Adaptive Landscapes of Resistance Genes Change as Antibiotic Concentrations Change

Epistasis-mediated compensatory evolution in a fitness landscape with adaptational tradeoffs

The evolutionary adaptation of an organism to a stressful environment often comes at the cost of reduced fitness. For example, resistance to antimicrobial drugs frequently reduces growth rate in the drug-free environment. This cost can be compensated without loss in resistance by mutations at secondary sites when the organism evolves again in the stress-free environment. Here, we analytically and numerically study evolution on a simple model fitness landscape to show that compensatory evolution can occur even in the presence of the stress and without the need for mutations at secondary sites. Fitness in the model depends on two phenotypes-the null-fitness defined as the fitness in the absence of stress, and the resistance level to the stress. Mutations universally exhibit antagonistic pleiotropy between the two phenotypes, that is they increase resistance while decreasing the null-fitness. Initial adaptation in this model occurs in a smooth region of the landscape with a rapid accumulation of stress resistance mutations and a concurrent decrease in the null-fitness. This is followed by a second, slower phase exhibiting partial recovery of the null-fitness. The second phase occurs on the rugged part of the landscape and involves the exchange of high-cost resistance mutations for low-cost ones. This process, which we call exchange compensation, is the result of changing epistatic interactions in the genotype as evolution progresses. The model provides general lessons about the tempo and mode of evolution under universal antagonistic pleiotropy with specific implications for drug resistance evolution.

Evolutionary druggability for low-dimensional fitness landscapes toward new metrics for antimicrobial applications

The term 'druggability' describes the molecular properties of drugs or targets in pharmacological interventions and is commonly used in work involving drug development for clinical applications. There are no current analogues for this notion that quantify the drug-target interaction with respect to a given target variant's sensitivity across a breadth of drugs in a panel, or a given drug's range of effectiveness across alleles of a target protein. Using data from low-dimensional empirical fitness landscapes composed of 16 β-lactamase alleles and 7 β-lactam drugs, we introduce two metrics that capture (i) the average susceptibility of an allelic variant of a drug target to any available drug in a given panel ('variant vulnerability'), and (ii) the average applicability of a drug (or mixture) across allelic variants of a drug target ('drug applicability'). Finally, we (iii) disentangle the quality and magnitude of interactions between loci in the drug target and the seven drug environments in terms of their mutation by mutation by environment (G x G x E) interactions, offering mechanistic insight into the variant variability and drug applicability metrics. Summarizing, we propose that our framework can be applied to other datasets and pathogen-drug systems to understand which pathogen variants in a clinical setting are the most concerning (low variant vulnerability), and which drugs in a panel are most likely to be effective in an infection defined by standing genetic variation in the pathogen drug target (high drug applicability).

Genotype sampling for deep-learning assisted experimental mapping of a combinatorially complete fitness landscape

MOTIVATION

Experimental characterization of fitness landscapes, which map genotypes onto fitness, is important for both evolutionary biology and protein engineering. It faces a fundamental obstacle in the astronomical number of genotypes whose fitness needs to be measured for any one protein. Deep learning may help to predict the fitness of many genotypes from a smaller neural network training sample of genotypes with experimentally measured fitness. Here I use a recently published experimentally mapped fitness landscape of more than 260 000 protein genotypes to ask how such sampling is best performed.

RESULTS

I show that multilayer perceptrons, recurrent neural networks, convolutional networks, and transformers, can explain more than 90% of fitness variance in the data. In addition, 90% of this performance is reached with a training sample comprising merely ≈103 sequences. Generalization to unseen test data is best when training data is sampled randomly and uniformly, or sampled to minimize the number of synonymous sequences. In contrast, sampling to maximize sequence diversity or codon usage bias reduces performance substantially. These observations hold for more than one network architecture. Simple sampling strategies may perform best when training deep learning neural networks to map fitness landscapes from experimental data.

AVAILABILITY AND IMPLEMENTATION

The fitness landscape data analyzed here is publicly available as described previously (Papkou et al. 2023). All code used to analyze this landscape is publicly available at https://github.com/andreas-wagner-uzh/fitness_landscape_sampling.

Understanding the effects of sub-inhibitory antibiotic concentrations on the development of β-lactamase resistance based on quantile regression analysis

AIMS

Quantile regression is an alternate type of regression analysis that has been shown to have numerous advantages over standard linear regression. Unlike linear regression, which uses the mean to fit a linear model, quantile regression uses a data set's quantiles (or percentiles), which leads to a more comprehensive analysis of the data. However, while relatively common in other scientific fields such as economic and environmental modeling, it is infrequently used to understand biological and microbiological systems.

METHODS AND RESULTS

We analyzed a set of bacterial growth rates using quantile regression analysis to better understand the effects of antibiotics on bacterial fitness. Using a bacterial model system containing 16 variant genotypes of the TEM β-lactamase enzyme, we compared our quantile regression analysis to a previously published study that uses the Tukey's range test, or Tukey honestly significantly difference (HSD) test. We find that trends in the distribution of bacterial growth rate data, as viewed through the lens of quantile regression, can distinguish between novel genotypes and ones that have been clinically isolated from patients. Quantile regression also identified certain combinations of genotypes and antibiotics that resulted in bacterial populations growing faster as the antibiotic concentration increased-the opposite of what was expected. These analyses can provide new insights into the relationships between enzymatic efficacy and antibiotic concentration.

CONCLUSIONS

Quantile regression analysis enhances our understanding of the impacts of sublethal antibiotic concentrations on enzymatic (TEM β-lactamase) efficacy and bacterial fitness. We illustrate that quantile regression analysis can link patterns in growth rates with clinically relevant mutations and provides an understanding of how increasing sub-lethal antibiotic concentrations, like those found in our modern environment, can affect bacterial growth rates, and provide insight into the genetic basis for varied resistance.

Epistasis and pleiotropy shape biophysical protein subspaces associated with drug resistance

Protein space is a rich analogy for genotype-phenotype maps, where amino acid sequence is organized into a high-dimensional space that highlights the connectivity between protein variants. It is a useful abstraction for understanding the process of evolution, and for efforts to engineer proteins towards desirable phenotypes. Few mentions of protein space consider how protein phenotypes can be described in terms of their biophysical components, nor do they rigorously interrogate how forces like epistasis-describing the nonlinear interaction between mutations and their phenotypic consequences-manifest across these components. In this study, we deconstruct a low-dimensional protein space of a bacterial enzyme (dihydrofolate reductase; DHFR) into "subspaces" corresponding to a set of kinetic and thermodynamic traits [k_{cat}, K_{M}, K_{i}, and T_{m} (melting temperature)]. We then examine how combinations of three mutations (eight alleles in total) display pleiotropy, or unique effects on individual subspace traits. We examine protein spaces across three orthologous DHFR enzymes (Escherichia coli, Listeria grayi, and Chlamydia muridarum), adding a genotypic context dimension through which epistasis occurs across subspaces. In doing so, we reveal that protein space is a deceptively complex notion, and that future applications to bioengineering should consider how interactions between amino acid substitutions manifest across different phenotypic subspaces.

Evolutionary druggability: leveraging low-dimensional fitness landscapes towards new metrics for antimicrobial applications

The term "druggability" describes the molecular properties of drugs or targets in pharmacological interventions and is commonly used in work involving drug development for clinical applications. There are no current analogues for this notion that quantify the drug-target interaction with respect to a given target variant's sensitivity across a breadth of drugs in a panel, or a given drug's range of effectiveness across alleles of a target protein. Using data from low-dimensional empirical fitness landscapes composed of 16 β-lactamase alleles and seven β-lactam drugs, we introduce two metrics that capture (i) the average susceptibility of an allelic variant of a drug target to any available drug in a given panel ("variant vulnerability"), and (ii) the average applicability of a drug (or mixture) across allelic variants of a drug target ("drug applicability"). Finally, we (iii) disentangle the quality and magnitude of interactions between loci in the drug target and the seven drug environments in terms of their mutation by mutation by environment (G × G × E) interactions, offering mechanistic insight into the variant variability and drug applicability metrics. Summarizing, we propose that our framework can be applied to other datasets and pathogen-drug systems to understand which pathogen variants in a clinical setting are the most concerning (low variant vulnerability), and which drugs in a panel are most likely to be effective in an infection defined by standing genetic variation in the pathogen drug target (high drug applicability).

Epistasis decreases with increasing antibiotic pressure but not temperature

Predicting mutational effects is essential for the control of antibiotic resistance (ABR). Predictions are difficult when there are strong genotype-by-environment (G × E), gene-by-gene (G × G or epistatic) or gene-by-gene-by-environment (G × G × E) interactions. We quantified G × G × E effects in Escherichia coli across environmental gradients. We created intergenic fitness landscapes using gene knock-outs and single-nucleotide ABR mutations previously identified to vary in the extent of G × E effects in our environments of interest. Then, we measured competitive fitness across a complete combinatorial set of temperature and antibiotic dosage gradients. In this way, we assessed the predictability of 15 fitness landscapes across 12 different but related environments. We found G × G interactions and rugged fitness landscapes in the absence of antibiotic, but as antibiotic concentration increased, the fitness effects of ABR genotypes quickly overshadowed those of gene knock-outs, and the landscapes became smoother. Our work reiterates that some single mutants, like those conferring resistance or susceptibility to antibiotics, have consistent effects across genetic backgrounds in stressful environments. Thus, although epistasis may reduce the predictability of evolution in benign environments, evolution may be more predictable in adverse environments. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

The Fitness of Beta-Lactamase Mutants Depends Nonlinearly on Resistance Level at Sublethal Antibiotic Concentrations

Adaptive evolutionary processes are constrained by the availability of mutations which cause a fitness benefit and together make up the fitness landscape, which maps genotype space onto fitness under specified conditions. Experimentally derived fitness landscapes have demonstrated a predictability to evolution by identifying limited "mutational routes" that evolution by natural selection may take between low and high-fitness genotypes. However, such studies often utilize indirect measures to determine fitness. We estimated the competitive fitness of mutants relative to all single-mutation neighbors to describe the fitness landscape of three mutations in a β-lactamase enzyme. Fitness assays were performed at sublethal concentrations of the antibiotic cefotaxime in a structured and unstructured environment. In the unstructured environment, the antibiotic selected for higher-resistance types-but with an equivalent fitness for a subset of mutants, despite substantial variation in resistance-resulting in a stratified fitness landscape. In contrast, in a structured environment with a low antibiotic concentration, antibiotic-susceptible genotypes had a relative fitness advantage, which was associated with antibiotic-induced filamentation. These results cast doubt that highly resistant genotypes have a unique selective advantage in environments with subinhibitory concentrations of antibiotics and demonstrate that direct fitness measures are required for meaningful predictions of the accessibility of evolutionary routes. IMPORTANCE The evolution of antibiotic-resistant bacterial populations underpins the ongoing antibiotic resistance crisis. We aim to understand how antibiotic-degrading enzymes can evolve to cause increased resistance, how this process is constrained, and whether it can be predictable. To this end, competition experiments were performed with a combinatorially complete set of mutants of a β-lactamase gene subject to subinhibitory concentrations of the antibiotic cefotaxime. While some mutations confer on their hosts high resistance to cefotaxime, in competition these mutations do not always confer a selective advantage. Specifically, high-resistance mutants had equivalent fitnesses despite different resistance levels and even had selective disadvantages under conditions involving spatial structure. Together, our findings suggest that the relationship between resistance level and fitness at subinhibitory concentrations is complex; predicting the evolution of antibiotic resistance requires knowledge of the conditions that select for resistant genotypes and the selective advantage evolved types have over their predecessors.

Epistasis meets pleiotropy in shaping biophysical protein subspaces associated with antimicrobial resistance

Protein space is a rich analogy for genotype-phenotype maps, where amino acid sequence is organized into a high-dimensional space that highlights the connectivity between protein variants. It is a useful abstraction for understanding the process of evolution, and for efforts to engineer proteins towards desirable phenotypes. Few framings of protein space consider how higher-level protein phenotypes can be described in terms of their biophysical dimensions, nor do they rigorously interrogate how forces like epistasis-describing the nonlinear interaction between mutations and their phenotypic consequences-manifest across these dimensions. In this study, we deconstruct a low-dimensional protein space of a bacterial enzyme (dihydrofolate reductase; DHFR) into "subspaces" corresponding to a set of kinetic and thermodynamic traits [(kcat, KM, Ki, and Tm (melting temperature)]. We then examine how three mutations (eight alleles in total) display pleiotropy in their interactions across these subspaces. We extend this approach to examine protein spaces across three orthologous DHFR enzymes (Escherichia coli, Listeria grayi, and Chlamydia muridarum), adding a genotypic context dimension through which epistasis occurs across subspaces. In doing so, we reveal that protein space is a deceptively complex notion, and that the process of protein evolution and engineering should consider how interactions between amino acid substitutions manifest across different phenotypic subspaces.

The population genetics of collateral resistance and sensitivity

Resistance mutations against one drug can elicit collateral sensitivity against other drugs. Multi-drug treatments exploiting such trade-offs can help slow down the evolution of resistance. However, if mutations with diverse collateral effects are available, a treated population may evolve either collateral sensitivity or collateral resistance. How to design treatments robust to such uncertainty is unclear. We show that many resistance mutations in Escherichia coli against various antibiotics indeed have diverse collateral effects. We propose to characterize such diversity with a joint distribution of fitness effects (JDFE) and develop a theory for describing and predicting collateral evolution based on simple statistics of the JDFE. We show how to robustly rank drug pairs to minimize the risk of collateral resistance and how to estimate JDFEs. In addition to practical applications, these results have implications for our understanding of evolution in variable environments.

Drugs known as antibiotics are the main treatment for most serious infections caused by bacteria. However, many bacteria are acquiring genetic mutations that make them resistant to the effects of one or more types of antibiotics, making them harder to eliminate.

One way to tackle drug-resistant bacteria is to develop new types of antibiotics; however, in recent years, the rate at which new antibiotics have become available has been dwindling. Using two or more existing drugs, one after another, can also be an effective way to eliminate resistant bacteria. The success of any such ‘multi-drug’ treatment lies in being able to predict whether mutations that make the bacteria resistant to one drug simultaneously make it sensitive to another, a phenomenon known as collateral sensitivity.

Different resistance mutations may have different collateral effects: some may increase the bacteria’s sensitivity to the second drug, while others might make the bacteria more resistant. However, it is currently unclear how to design robust multi-drug treatments that take this diversity of collateral effects into account.

Here, Ardell and Kryazhimskiy used a concept called JDFE (short for the joint distribution of fitness effects) to describe the diversity of collateral effects in a population of bacteria exposed to a single drug. This information was then used to mathematically model how collateral effects evolved in the population over time.

Ardell and Kryazhimskiy showed that this approach can predict how likely a population is to become collaterally sensitive or collaterally resistant to a second antibiotic. Drug pairs can then be ranked according to the risk of collateral resistance emerging, so long as information on the variety of resistance mutations available to the bacteria are included in the model.

Each year, more than 700,000 people die from infections caused by bacteria that are resistant to one or more antibiotics. The findings of Ardell and Kryazhimskiy may eventually help clinicians design multi-drug treatments that effectively eliminate bacterial infections and help to prevent more bacteria from evolving resistance to antibiotics. However, to achieve this goal, more research is needed to fully understand the range collateral effects caused by resistance mutations.

Hidden role of mutations in the evolutionary process

Mutations not only alter allele frequencies in a genetic pool but may also determine the fate of an evolutionary process. Here we study which allele fixes in a one-step, one-way model including the wild type and two adaptive mutations. We study the effect of the four basic evolutionary mechanisms-genetic drift, natural selection, mutation, and gene flow-on mutant fixation and its kinetics. Determining which allele is more likely to fix is not simply a question of comparing fitnesses and mutation rates. For instance, if the allele of interest is less fit than the other, then not only must it have a greater mutation rate, but also its mutation rate must exceed a specific threshold for it to prevail. We find exact expressions for such conditions. Our conclusions are based on the mathematical description of two extreme but important regimes, as well as on simulations.

Resistance elicited by sub-lethal concentrations of ampicillin is partially mediated by quorum sensing in Pseudomonas aeruginosa

The rapid increase of antibiotic resistance is a serious challenge around the world. Antibiotics are present in various environments at sub-lethal concentrations, but how resistance emerges under sub-lethal conditions is not fully clear. In this study, we evolved Pseudomonas aeruginosa PAO1 under sub-lethal conditions, in the presence of either 15-30 μg/mL or 150-300 μg/mL of ampicillin. We found a ~ 5-6 fold increase in the minimum inhibitory concentration (MIC) among evolved isolates exposed to 15-30 μg/mL of ampicillin, and more than a 19-fold of increase in 150-300 μg/mL of ampicillin exposure. DNA sequencing revealed that mpl and ampD were frequently mutated in these resistant strains. We performed a transcriptome analysis of deletion mutations of mpl or ampD, compared to PAO1. Both showed a two-fold increase in expression of quorum sensing (QS) genes including lasR and rhlI/R; the heightened expression was positively correlated with the expression of the ampicillin resistance gene ampC. We queried if quorum sensing contributes to the increase in the ampicillin MIC. After adding the quorum quencher acylase I, the growth yield both decreased by roughly 50% for Δmpl in 2000 μg/mL of ampicillin and ΔampD in 4000 μg/mL of ampicillin. Addition of the QS signals into synthase mutants restored the higher MIC, but only for the rhlI/R circuit. This study highlights the involvement of QS in antibiotic resistance evolution, and shows the multifactorial contributors to the observed phenotypes.

The Antimicrobial Resistance Crisis: An Inadvertent, Unfortunate but Nevertheless Informative Experiment in Evolutionary Biology

The global rise of antimicrobial resistance (AMR) phenotypes is an exemplar for rapid evolutionary response. Resistance arises as a consequence of humanity’s widespread and largely indiscriminate use of antimicrobial compounds. However, some features of this crisis remain perplexing. The remarkably widespread and rapid rise of diverse, novel and effective resistance phenotypes is in stark contrast to the apparent paucity of antimicrobial producers in the global microbiota. From the viewpoint of evolutionary theory, it should be possible to use selection coefficients to examine these phenomena. In this work we introduce an elaboration on the selection coefficient s termed selective efficiency, considering the genetic, metabolic, ecological and evolutionary impacts that accompany selective phenotypes. We then demonstrate the utility of the selective efficiency concept using AMR and antimicrobial production phenotypes as ‘worked examples’ of the concept. In accomplishing this objective, we also put forward cogent hypotheses to explain currently puzzling aspects of the AMR crisis. Finally, we extend the selective efficiency concept into a consideration of the ongoing management of the AMR crisis.

Relevance of Higher-Order Epistasis in Drug Resistance

We studied five chemically distinct but related 1,3,5-triazine antifolates with regard to their effects on growth of a set of mutants in dihydrofolate reductase. The mutants comprise a combinatorially complete data set of all 16 possible combinations of four amino acid replacements associated with resistance to pyrimethamine in the malaria parasite Plasmodium falciparum. Pyrimethamine was a mainstay medication for malaria for many years, and it is still in use in intermittent treatment during pregnancy or as a partner drug in artemisinin combination therapy. Our goal was to investigate the extent to which the alleles yield similar adaptive topographies and patterns of epistasis across chemically related drugs. We find that the adaptive topographies are indeed similar with the same or closely related alleles being fixed in computer simulations of stepwise evolution. For all but one of the drugs the topography features at least one suboptimal fitness peak. Our data are consistent with earlier results indicating that third order and higher epistatic interactions appear to contribute only modestly to the overall adaptive topography, and they are largely conserved. In regard to drug development, our data suggest that higher-order interactions are likely to be of little value as an advisory tool in the choice of lead compounds.

A two-step PCR assembly for construction of gene variants across large mutational distances

Construction of empirical fitness landscapes has transformed our understanding of genotype-phenotype relationships across genes. However, most empirical fitness landscapes have been constrained to the local genotype neighbourhood of a gene primarily due to our limited ability to systematically construct genotypes that differ by a large number of mutations. Although a few methods have been proposed in the literature, these techniques are complex owing to several steps of construction or contain a large number of amplification cycles that increase chances of non-specific mutations. A few other described methods require amplification of the whole vector, thereby increasing the chances of vector backbone mutations that can have unintended consequences for study of fitness landscapes. Thus, this has substantially constrained us from traversing large mutational distances in the genotype network, thereby limiting our understanding of the interactions between multiple mutations and the role these interactions play in evolution of novel phenotypes. In the current work, we present a simple but powerful approach that allows us to systematically and accurately construct gene variants at large mutational distances. Our approach relies on building-up small fragments containing targeted mutations in the first step followed by assembly of these fragments into the complete gene fragment by polymerase chain reaction (PCR). We demonstrate the utility of our approach by constructing variants that differ by up to 11 mutations in a model gene. Our work thus provides an accurate method for construction of multi-mutant variants of genes and therefore will transform the studies of empirical fitness landscapes by enabling exploration of genotypes that are far away from a starting genotype.

Adaptive Processes Change as Multiple Functions Evolve

Epistasis influences the gene-environment interactions that shape bacterial fitness through antibiotic exposure, which can ultimately affect the availability of certain resistance phenotypes to bacteria. The substitutions present within bla _TEM-50 confer both cephalosporin and β-lactamase inhibitor resistance. We wanted to compare the evolution of bla _TEM-50 with that of another variant, bla _TEM-85, which differs in that bla _TEM-85 contains only substitutions that contribute to cephalosporin resistance. Differences between the landscapes and epistatic interactions of these TEM variants are important for understanding their separate evolutionary responses to antibiotics. We hypothesized the substitutions within bla _TEM-50 would result in more epistatic interactions than for bla _TEM-85 As expected, we found more epistatic interactions between the substitutions present in bla _TEM-50 than in bla _TEM-85 Our results suggest that selection from many cephalosporins is required to achieve the full potential resistance to cephalosporins but that a single β-lactam and inhibitor combination will drive evolution of the full potential resistance phenotype. Surprisingly, we also found significantly positive increases in growth rates as antibiotic concentration increased for some of the strains expressing bla _TEM-85 precursor genotypes but not the bla _TEM-50 variants. This result further suggests that additive interactions more effectively optimize phenotypes than epistatic interactions, which means that exposure to numerous cephalosporins actually increases the ability of a TEM enzyme to confer resistance to any single cephalosporin.

Predicting antimicrobial resistance using conserved genes

A growing number of studies are using machine learning models to accurately predict antimicrobial resistance (AMR) phenotypes from bacterial sequence data. Although these studies are showing promise, the models are typically trained using features derived from comprehensive sets of AMR genes or whole genome sequences and may not be suitable for use when genomes are incomplete. In this study, we explore the possibility of predicting AMR phenotypes using incomplete genome sequence data. Models were built from small sets of randomly-selected core genes after removing the AMR genes. For Klebsiella pneumoniae, Mycobacterium tuberculosis, Salmonella enterica, and Staphylococcus aureus, we report that it is possible to classify susceptible and resistant phenotypes with average F1 scores ranging from 0.80-0.89 with as few as 100 conserved non-AMR genes, with very major error rates ranging from 0.11-0.23 and major error rates ranging from 0.10-0.20. Models built from core genes have predictive power in cases where the primary AMR mechanisms result from SNPs or horizontal gene transfer. By randomly sampling non-overlapping sets of core genes, we show that F1 scores and error rates are stable and have little variance between replicates. Although these small core gene models have lower accuracies and higher error rates than models built from the corresponding assembled genomes, the results suggest that sufficient variation exists in the core non-AMR genes of a species for predicting AMR phenotypes.

Candida albicans Genetic Background Influences Mean and Heterogeneity of Drug Responses and Genome Stability during Evolution in Fluconazole

Antimicrobial resistance is an evolutionary phenomenon with clinical implications. We tested how replicates from diverse strains of Candida albicans, a prevalent human fungal pathogen, evolve in the commonly prescribed antifungal drug fluconazole. Replicates on average increased in fitness in the level of drug they were evolved to, with the least fit parental strains improving the most. Very few replicates increased resistance above the drug level they were evolved in. Notably, many replicates increased in genome size and changed in drug tolerance (a drug response where a subpopulation of cells grow slowly in high levels of drug), and variability among replicates in fitness, tolerance, and genome size was higher in strains that initially were more sensitive to the drug. Genetic background influenced the average degree of adaptation and the evolved variability of many phenotypes, highlighting that different strains from the same species may respond and adapt very differently during adaptation.

The importance of within-species diversity in determining the evolutionary potential of a population to evolve drug resistance or tolerance is not well understood, including in eukaryotic pathogens. To examine the influence of genetic background, we evolved replicates of 20 different clinical isolates of Candida albicans, a human fungal pathogen, in fluconazole, the commonly used antifungal drug. The isolates hailed from the major C. albicans clades and had different initial levels of drug resistance and tolerance to the drug. The majority of replicates rapidly increased in fitness in the evolutionary environment, with the degree of improvement inversely correlated with parental strain fitness in the drug. Improvement was largely restricted to up to the evolutionary level of drug: only 4% of the evolved replicates increased resistance (MIC) above the evolutionary level of drug. Prevalent changes were altered levels of drug tolerance (slow growth of a subpopulation of cells at drug concentrations above the MIC) and increased diversity of genome size. The prevalence and predominant direction of these changes differed in a strain-specific manner, but neither correlated directly with parental fitness or improvement in fitness. Rather, low parental strain fitness was correlated with high levels of heterogeneity in fitness, tolerance, and genome size among evolved replicates. Thus, parental strain background is an important determinant in mean improvement to the evolutionary environment as well as the diversity of evolved phenotypes, and the range of possible responses of a pathogen to an antimicrobial drug cannot be captured by in-depth study of a single strain background.

IMPORTANCE Antimicrobial resistance is an evolutionary phenomenon with clinical implications. We tested how replicates from diverse strains of Candida albicans, a prevalent human fungal pathogen, evolve in the commonly prescribed antifungal drug fluconazole. Replicates on average increased in fitness in the level of drug they were evolved to, with the least fit parental strains improving the most. Very few replicates increased resistance above the drug level they were evolved in. Notably, many replicates increased in genome size and changed in drug tolerance (a drug response where a subpopulation of cells grow slowly in high levels of drug), and variability among replicates in fitness, tolerance, and genome size was higher in strains that initially were more sensitive to the drug. Genetic background influenced the average degree of adaptation and the evolved variability of many phenotypes, highlighting that different strains from the same species may respond and adapt very differently during adaptation.

Predictable properties of fitness landscapes induced by adaptational tradeoffs

Fitness effects of mutations depend on environmental parameters. For example, mutations that increase fitness of bacteria at high antibiotic concentration often decrease fitness in the absence of antibiotic, exemplifying a tradeoff between adaptation to environmental extremes. We develop a mathematical model for fitness landscapes generated by such tradeoffs, based on experiments that determine the antibiotic dose-response curves of Escherichia coli strains, and previous observations on antibiotic resistance mutations. Our model generates a succession of landscapes with predictable properties as antibiotic concentration is varied. The landscape is nearly smooth at low and high concentrations, but the tradeoff induces a high ruggedness at intermediate antibiotic concentrations. Despite this high ruggedness, however, all the fitness maxima in the landscapes are evolutionarily accessible from the wild type. This implies that selection for antibiotic resistance in multiple mutational steps is relatively facile despite the complexity of the underlying landscape.

Growth rate assays reveal fitness consequences of β-lactamases

Antibiotic resistance is a powerful model for studying evolutionary biology and population genetics. For the purpose of these evolutionary studies, fitness data have been approximated through susceptibility testing methods designed for clinical use in providing appropriate antibiotic therapies. An alternative approach for measuring fitness of microbes has experienced growing popularity: growth rates are a highly sensitive approach for measuring the fitness effects of antibiotics and resistance genes, but they differ from susceptibility testing in that a single concentration of antibiotic is used for the assay. Here we show that despite this key difference, the results of growth rates correlate well with clinical determination of resistance by minimum inhibitory concentration (MIC), while providing the sensitivity required for direct input as fitness values into mathematical models. This means that growth rates at a single sublethal inhibitory concentration can help us understand the fitness effects that result in clinical antibiotic resistance. By measuring the growth rates of sequenced clinical isolates obtained from Dignity Health Mercy Medical Center, we detected the fitness effects of individual resistance genes on bacteria exposed to different antibiotics. In our study, the CTX-M-15 gene conferred the highest fitness in assays with cephalosporins. These results show the strong fitness benefit conferred by CTX-M-15.

Catchment-scale export of antibiotic resistance genes and bacteria from an agricultural watershed in central Iowa

Antibiotics are administered to livestock in animal feeding operations (AFOs) for the control, prevention, and treatment of disease. Manure from antibiotic treated livestock contains unmetabolized antibiotics that provide selective pressure on bacteria, facilitating the expression of anti-microbial resistance (AMR). Manure application on row crops is an agronomic practice used by growers to meet crop nutrient needs; however, it can be a source of AMR to the soil and water environment. This study in central Iowa aims to directly compare AMR indicators in outlet runoff from two adjacent (221 to 229 ha) manured and non-manured catchments (manure comparison), and among three catchments (600 to 804 ha) with manure influence, no known manure application (control), and urban influences (mixed land use comparison). Monitored AMR indicators included antibiotic resistance genes (ARGs) ermB, ermF (macrolide), tetA, tetM, tetO, tetW (tetracycline), sul1, sul2 (sulfonamide), aadA2 (aminoglycoside), vgaA, and vgaB (pleuromutilin), and tylosin and tetracycline resistant enterococci bacteria. Results of the manure comparison showed significantly higher (p<0.05) tetracycline and tylosin resistant bacteria from the catchment with manure application in 2017, but no differences in 2018, possibly due to changes in antibiotic use resulting from the Veterinary Feed Directive. Moreover, the ARG analysis indicated a larger diversity of ARGs at the manure amended catchment. The mixed land use comparison showed the manure amended catchment had significantly higher (p<0.05) tetracycline resistant bacteria in 2017 and significantly higher tylosin resistant bacteria in 2017 and 2018 than the urban influenced catchment. The urban influenced catchment had significantly higher ermB concentrations in both sampling years, however the manure applied catchment runoff consisted of higher relative abundance of total ARGs. Additionally, both catchments showed higher AMR indicators compared to the control catchment. This study identifies four ARGs that might be specific to AMR as a result of agricultural sources (tetM, tetW, sul1, sul2) and optimal for use in watershed scale monitoring studies for tracking resistance in the environment.

Diffusion-driven enhancement of the antibiotic resistance selection window

The current crisis of antimicrobial resistance in clinically relevant pathogens has highlighted our limited understanding of the ecological and evolutionary forces that drive drug resistance adaptation. For instance, although human tissues are highly heterogeneous, most of our mechanistic understanding about antibiotic resistance evolution is based on constant and well-mixed environmental conditions. A consequence of considering spatial heterogeneity is that, even if antibiotics are prescribed at high dosages, the penetration of drug molecules through tissues inevitably produces antibiotic gradients, exposing bacterial populations to a range of selective pressures and generating a dynamic fitness landscape that changes in space and time. In this paper, we will use a combination of mathematical modelling and computer simulations to study the population dynamics of susceptible and resistant strains competing for resources in a network of micro-environments with varying degrees of connectivity. Our main result is that highly connected environments increase diffusion of drug molecules, enabling resistant phenotypes to colonize a larger number of spatial locations. We validated this theoretical result by culturing fluorescently labelled Escherichia coli in 3D-printed devices that allow us to control the rate of diffusion of antibiotics between neighbouring compartments and quantify the spatio-temporal distribution of resistant and susceptible bacterial cells.

Growth kinetics in Candida spp.: Differences between species and potential impact on antifungal susceptibility testing as described by the EUCAST

We studied the growth kinetic parameters of clinically relevant Candida species to verify the differences between species following the incubation and medium conditions recommended by the EUCAST. We analyzed 705 susceptible Candida spp. from patients with candidemia and Candida glabrata isolates resistant to echinocandins or fluconazole (n = 38) and calculated the average growth rate, maximum peak, time to maximum rate, and lag phase. We also examined inter- and intra-species differences, as well as the percentage of isolates reaching an optical density of 0.2 over time. Interspecies differences in growth phases and kinetic parameters were found. C. glabrata was the fastest growing species and the lag phase of C. parapsilosis was longer than that of the other species considered in this study. Strain-to-strain variations were found between species. A positive correlation between the average growth rate and maximum peak was determined. Echinocandin-resistant C. glabrata isolates had significantly lower average growth rate but higher time to maximum rate in comparison to wild-type C. glabrata isolates. Incubation periods of 12-15 hours allowed reaching the 0.2 optical density threshold in 100% of C. glabrata, C. tropicalis, and C. krusei isolates. We show differences in kinetic parameters between Candida spp. C. glabrata was the fastest growing species and C. parapsilosis showed the longest lag phase. Resistance to echinocandins may affect the growth kinetic curve. Speeding up antifungal susceptibility results could be possible for some isolates, particularly C. glabrata, C. tropicalis, and C. krusei.

Adaptive Landscapes in the Age of Synthetic Biology

For nearly a century adaptive landscapes have provided overviews of the evolutionary process and yet they remain metaphors. We redefine adaptive landscapes in terms of biological processes rather than descriptive phenomenology. We focus on the underlying mechanisms that generate emergent properties such as epistasis, dominance, trade-offs and adaptive peaks. We illustrate the utility of landscapes in predicting the course of adaptation and the distribution of fitness effects. We abandon aged arguments concerning landscape ruggedness in favor of empirically determining landscape architecture. In so doing, we transform the landscape metaphor into a scientific framework within which causal hypotheses can be tested.

Effects of a previously selected antibiotic resistance on mutations acquired during development of a second resistance in Escherichia coli

Background

The effect of mutations conferring antibiotic resistance can depend on the genetic background. To determine if a previously de novo acquired antibiotic resistance influences the adaptation to a second antibiotic, antibiotic resistance was selected for by exposure to stepwise increasing sublethal levels of amoxicillin, enrofloxacin, kanamycin, or tetracycline. E. coli populations adapted to either a single or two antibiotics sequentially were characterized using whole genome population sequencing and MIC measurements.

Results

In a wild-type background, adaptation to any of the antibiotics resulted in the appearance of well-known mutations, as well as a number of mutated genes not known to be associated with antibiotic resistance. Development of a second resistance in a strain with an earlier acquired resistance to a different antibiotic did not always result in the appearance of all mutations associated with resistance in a wild-type background. In general, a more varied set of mutations was acquired during secondary adaptation. The ability of E. coli to maintain the first resistance during this process depended on the combination of antibiotics used. The maintenance of mutations associated with resistance to the first antibiotic did not always predict the residual MIC for that compound.

Conclusions

In general, the data presented here indicate that adaptation to each antibiotic is unique and independent. The mutational trajectories available in already resistant cells appear more varied than in wild-type cells, indicating that the genetic background of E. coli influences resistance development. The observed mutations cannot always fully explain the resistance pattern observed, indicating a crucial role for adaptation on the gene expression level in de novo acquisition of antibiotic resistance.

Dose-dependent impact of oxytetracycline on the veal calf microbiome and resistome

BACKGROUND

Antibiotic therapy is commonly used in animal agriculture. Antibiotics excreted by the animals can contaminate farming environments, resulting in long term exposure of animals to sub-inhibitory levels of antibiotics. Little is known on the effect of this exposure on antibiotic resistance. In this study, we aimed to investigate the long term effects of sub-inhibitory levels of antibiotics on the gut microbiota composition and resistome of veal calves in vivo. Forty-two veal calves were randomly assigned to three groups. The first group (OTC-high) received therapeutic oral dosages of 1 g oxytetracycline (OTC), twice per day, during 5 days. The second group (OTC-low) received an oral dose of OTC of 100-200 μg per day during 7 weeks, mimicking animal exposure to environmental contamination. The third group (CTR) did not receive OTC, serving as unexposed control. Antibiotic residue levels were determined over time. The temporal effects on the gut microbiota and antibiotic resistance gene abundance was analysed by metagenomic sequencing.

RESULTS

In the therapeutic group, OTC levels exceeded MIC values. The low group remained at sub-inhibitory levels. The control group did not reach any significant OTC levels. 16S rRNA gene-based analysis revealed significant changes in the calf gut microbiota. Time-related changes accounted for most of the variation in the sequence data. Therapeutic application of OTC had transient effect, significantly impacting gut microbiota composition between day 0 and day 2. By metagenomic sequence analysis we identified six antibiotic resistance genes representing three gene classes (tetM, floR and mel) that differed in relative abundance between any of the intervention groups and the control. qPCR was used to validate observations made by metagenomic sequencing, revealing a peak of tetM abundance at day 28-35 in the OTC-high group. No increase in resistance genes abundance was seen in the OTC-low group.

CONCLUSIONS

Under the conditions tested, sub-therapeutic administration of OTC did not result in increased tetM resistance levels as observed in the therapeutic group.

Fitness Penalties in the Evolution of Fungicide Resistance

The evolution of resistance poses an ongoing threat to crop protection. Fungicide resistance provides a selective advantage under fungicide selection, but resistance-conferring mutations may also result in fitness penalties, resulting in an evolutionary trade-off. These penalties may result from the functional constraints of an evolving target site or from the resource allocation costs of overexpression or active transport. The extent to which such fitness penalties are present has important implications for resistance management strategies, determining whether resistance persists or declines between treatments, and for resistance risk assessments for new modes of action. Experimental results have proven variable, depending on factors such as temperature, nutrient status, osmotic or oxidative stress, and pathogen life-cycle stage. Functional genetics tools allow pathogen genetic background to be controlled, but this in turn raises the question of epistatic interactions. Combining fitness penalties under various conditions into a field-realistic scenario poses an important future challenge.

Evolutionary constraints in fitness landscapes

In the last years, several genotypic fitness landscapes-combinations of a small number of mutations-have been experimentally resolved. To learn about the general properties of "real" fitness landscapes, it is key to characterize these experimental landscapes via simple measures of their structure, related to evolutionary features. Some of the most relevant measures are based on the selectively acessible paths and their properties. In this paper, we present some measures of evolutionary constraints based on (i) the similarity between accessible paths and (ii) the abundance and characteristics of "chains" of obligatory mutations, that are paths going through genotypes with a single fitter neighbor. These measures have a clear evolutionary interpretation. Furthermore, we show that chains are only weakly correlated to classical measures of epistasis. In fact, some of these measures of constraint are non-monotonic in the amount of epistatic interactions, but have instead a maximum for intermediate values. Finally, we show how these measures shed light on evolutionary constraints and predictability in experimentally resolved landscapes.

Mapping Conformational Dynamics to Individual Steps in the TEM-1 β-Lactamase Catalytic Mechanism

Conformational dynamics are increasingly recognized as being essential for enzyme function. However, there is virtually no direct experimental evidence to support the notion that individual dynamic modes are required for specific catalytic processes, apart from the initial step of substrate binding. In this work, we use a unique approach based on millisecond hydrogen-deuterium exchange mass spectrometry to identify dynamic modes linked to individual catalytic processes in the antibiotic resistance enzyme TEM-1 β-lactamase. Using a "good" substrate (ampicillin), a poorly hydrolyzed substrate (cephalexin) and a covalent inhibitor (clavulanate), we are able to isolate dynamic modes that are specifically linked to substrate binding, productive lactam ring hydrolysis and deacylation. These discoveries are ultimately translated into specific targets for allosteric TEM-1 inhibitor development.

The Influence of Higher-Order Epistasis on Biological Fitness Landscape Topography

The effect of a mutation on the organism often depends on what other mutations are already present in its genome. Geneticists refer to such mutational interactions as epistasis. Pairwise epistatic effects have been recognized for over a century, and their evolutionary implications have received theoretical attention for nearly as long. However, pairwise epistatic interactions themselves can vary with genomic background. This is called higher-order epistasis, and its consequences for evolution are much less well understood. Here, we assess the influence that higher-order epistasis has on the topography of 16 published, biological fitness landscapes. We find that on average, their effects on fitness landscape declines with order, and suggest that notable exceptions to this trend may deserve experimental scrutiny. We conclude by highlighting opportunities for further theoretical and experimental work dissecting the influence that epistasis of all orders has on fitness landscape topography and on the efficiency of evolution by natural selection.

Clinical Evolution of New Delhi Metallo-β-Lactamase (NDM) Optimizes Resistance under Zn(II) Deprivation

Carbapenem-resistant Enterobacteriaceae (CRE) are rapidly spreading and taking a staggering toll on all health care systems, largely due to the dissemination of genes coding for potent carbapenemases. An important family of carbapenemases are the Zn(II)-dependent β-lactamases, known as metallo-β-lactamases (MBLs). Among them, the New Delhi metallo-β-lactamase (NDM) has experienced the fastest and widest geographical spread. While other clinically important MBLs are soluble periplasmic enzymes, NDMs are lipoproteins anchored to the outer membrane in Gram-negative bacteria. This unique cellular localization endows NDMs with enhanced stability upon the Zn(II) starvation elicited by the immune system response at the sites of infection. Since the first report of NDM-1, new allelic variants (16 in total) have been identified in clinical isolates differing by a limited number of substitutions. Here, we show that these variants have evolved by accumulating mutations that enhance their stability or the Zn(II) binding affinity in vivo, overriding the most common evolutionary pressure acting on catalytic efficiency. We identified the ubiquitous substitution M154L as responsible for improving the Zn(II) binding capabilities of the NDM variants. These results also reveal that Zn(II) deprivation imposes a strict constraint on the evolution of this MBL, overriding the most common pressures acting on catalytic performance, and shed light on possible inhibitory strategies.

Statistical Package for Growth Rates Made Easy

Growth rates are an important tool in microbiology because they provide high throughput fitness measurements. The release of GrowthRates, a program that uses the output of plate reader files to automatically calculate growth rates, has facilitated experimental procedures in many areas. However, many sources of variation within replicate growth rate data exist and can decrease data reliability. We have developed a new statistical package, CompareGrowthRates (CGR), to enhance the program GrowthRates and accurately measure variation in growth rate data sets. We define a metric, Variability-score (V-score), that can help determine if variation within a data set might result in false interpretations. CGR also uses the bootstrap method to determine the fraction of bootstrap replicates in which a strain will grow the fastest. We illustrate the usage of CGR with growth rate data sets similar to those in Mira, Meza, et al. (Adaptive landscapes of resistance genes change as antibiotic concentrations change. Mol Biol Evol. 32(10): 2707-2715). These statistical methods are compatible with the analytic methods described in Growth Rates Made Easy and can be used with any set of growth rate output from GrowthRates.

Local Fitness Landscapes Predict Yeast Evolutionary Dynamics in Directionally Changing Environments

The fitness landscape is a concept that is widely used for understanding and predicting evolutionary adaptation. The topography of the fitness landscape depends critically on the environment, with potentially far-reaching consequences for evolution under changing conditions. However, few studies have assessed directly how empirical fitness landscapes change across conditions, or validated the predicted consequences of such change. We previously evolved replicate yeast populations in the presence of either gradually increasing, or constant high, concentrations of the heavy metals cadmium (Cd), nickel (Ni), and zinc (Zn), and analyzed their phenotypic and genomic changes. Here, we reconstructed the local fitness landscapes underlying adaptation to each metal by deleting all repeatedly mutated genes both by themselves and in combination. Fitness assays revealed that the height, and/or shape, of each local fitness landscape changed considerably across metal concentrations, with distinct qualitative differences between unconditionally (Cd) and conditionally toxic metals (Ni and Zn). This change in topography had particularly crucial consequences in the case of Ni, where a substantial part of the individual mutational fitness effects changed in sign across concentrations. Based on the Ni landscape analyses, we made several predictions about which mutations had been selected when during the evolution experiment. Deep sequencing of population samples from different time points generally confirmed these predictions, demonstrating the power of landscape reconstruction analyses for understanding and ultimately predicting evolutionary dynamics, even under complex scenarios of environmental change.

History of antibiotic adaptation influences microbial evolutionary dynamics during subsequent treatment

Antibiotic regimens often include the sequential changing of drugs to limit the development and evolution of resistance of bacterial pathogens. It remains unclear how history of adaptation to one antibiotic can influence the resistance profiles when bacteria subsequently adapt to a different antibiotic. Here, we experimentally evolved Pseudomonas aeruginosa to six 2-drug sequences. We observed drug order-specific effects, whereby adaptation to the first drug can limit the rate of subsequent adaptation to the second drug, adaptation to the second drug can restore susceptibility to the first drug, or final resistance levels depend on the order of the 2-drug sequence. These findings demonstrate how resistance not only depends on the current drug regimen but also the history of past regimens. These order-specific effects may allow for rational forecasting of the evolutionary dynamics of bacteria given knowledge of past adaptations and provide support for the need to consider the history of past drug exposure when designing strategies to mitigate resistance and combat bacterial infections.

Epistasis and the Evolution of Antimicrobial Resistance

The fitness effects of a mutation can depend, sometimes dramatically, on genetic background; this phenomenon is often referred to as “epistasis.” Epistasis can have important practical consequences in the context of antimicrobial resistance (AMR). For example, genetic background plays an important role in determining the costs of resistance, and hence in whether resistance will persist in the absence of antibiotic pressure. Furthermore, interactions between resistance mutations can have important implications for the evolution of multi-drug resistance. I argue that there is a need to better characterize the extent and nature of epistasis for mutations and horizontally transferred elements conferring AMR, particularly in clinical contexts. Furthermore, I suggest that epistasis should be an important consideration in attempts to slow or limit the evolution of AMR.

Widespread Genetic Incompatibilities between First-Step Mutations during Parallel Adaptation of Saccharomyces cerevisiae to a Common Environment

Independently evolving populations may adapt to similar selection pressures via different genetic changes. The interactions between such changes, such as in a hybrid individual, can inform us about what course adaptation may follow and allow us to determine whether gene flow would be facilitated or hampered following secondary contact. We used Saccharomyces cerevisiae to measure the genetic interactions between first-step mutations that independently evolved in the same biosynthetic pathway following exposure to the fungicide nystatin. We found that genetic interactions are prevalent and predominantly negative, with the majority of mutations causing lower growth when combined in a double mutant than when alone as a single mutant (sign epistasis). The prevalence of sign epistasis is surprising given the small number of mutations tested and runs counter to expectations for mutations arising in a single biosynthetic pathway in the face of a simple selective pressure. Furthermore, in one third of pairwise interactions, the double mutant grew less well than either single mutant (reciprocal sign epistasis). The observation of reciprocal sign epistasis among these first adaptive mutations arising in the same genetic background indicates that partial postzygotic reproductive isolation could evolve rapidly between populations under similar selective pressures, even with only a single genetic change in each. The nature of the epistatic relationships was sensitive, however, to the level of drug stress in the assay conditions, as many double mutants became fitter than the single mutants at higher concentrations of nystatin. We discuss the implications of these results both for our understanding of epistatic interactions among beneficial mutations in the same biochemical pathway and for speciation.

Tackling Threats and Future Problems of Multidrug-Resistant Bacteria

With the advent of the antibiotic era, the overuse and inappropriate consumption and application of antibiotics have driven the rapid emergence of multidrug-resistant pathogens. Antimicrobial resistance increases the morbidity, mortality, length of hospitalization and healthcare costs. Among Gram-positive bacteria, Staphylococcus aureus (MRSA) and multidrug-resistant (MDR) Mycobacterium tuberculosis, and among the Gram-negative bacteria, extended-spectrum beta-lactamase (ESBLs)-producing bacteria have become a major global healthcare problem in the 21st century. The pressure to use antibiotics guarantees that the spread and prevalence of these as well as of future emerging multidrug-resistant pathogens will be a persistent phenomenon. The unfeasibility of reversing antimicrobial resistance back towards susceptibility and the critical need to treat bacterial infection in modern medicine have burdened researchers and pharmaceutical companies to develop new antimicrobials effective against these difficult-to-treat multidrug-resistant pathogens. However, it can be anticipated that antibiotic resistance will continue to develop more rapidly than new agents to treat these infections become available and a better understanding of the molecular, evolutionary and ecological mechanisms governing the spread of antibiotic resistance is needed. The only way to curb the current crisis of antimicrobial resistance will be to develop entirely novel strategies to fight these pathogens such as combining antimicrobial drugs with other agents that counteract and obstruct the antibiotic resistant mechanisms expressed by the pathogen. Furthermore, as many antibiotics are often inappropriately prescribed, a more personalized approach based on precise diagnosis tools will ensure that proper treatments can be promptly applied leading to more targeted and effective therapies. However, in more general terms, also the overall use and release of antibiotics in the environment needs to be better controlled.

Drug-target networks in aminoglycoside resistance: hierarchy of priority in structural drug design

Drug-target network analysis for advancing next-generation aminoglycoside therapies that combat antibiotic resistant infections.