Predicting mutational routes to new adaptive phenotypes

Prediction of mutation-induced protein stability changes based on the geometric representations learned by a self-supervised method

BACKGROUND

Thermostability is a fundamental property of proteins to maintain their biological functions. Predicting protein stability changes upon mutation is important for our understanding protein structure-function relationship, and is also of great interest in protein engineering and pharmaceutical design.

RESULTS

Here we present mutDDG-SSM, a deep learning-based framework that uses the geometric representations encoded in protein structure to predict the mutation-induced protein stability changes. mutDDG-SSM consists of two parts: a graph attention network-based protein structural feature extractor that is trained with a self-supervised learning scheme using large-scale high-resolution protein structures, and an eXtreme Gradient Boosting model-based stability change predictor with an advantage of alleviating overfitting problem. The performance of mutDDG-SSM was tested on several widely-used independent datasets. Then, myoglobin and p53 were used as case studies to illustrate the effectiveness of the model in predicting protein stability changes upon mutations. Our results show that mutDDG-SSM achieved high performance in estimating the effects of mutations on protein stability. In addition, mutDDG-SSM exhibited good unbiasedness, where the prediction accuracy on the inverse mutations is as well as that on the direct mutations.

CONCLUSION

Meaningful features can be extracted from our pre-trained model to build downstream tasks and our model may serve as a valuable tool for protein engineering and drug design.

Genomic sequencing reveals convergent adaptation during experimental evolution in two budding yeast species

Convergent evolution is central in the origins of multicellularity. Identifying the basis for convergent multicellular evolution is challenging because of the diverse evolutionary origins and environments involved. Haploid Kluyveromyces lactis populations evolve multicellularity during selection for increased settling in liquid media. Strong genomic and phenotypic convergence is observed between K. lactis and previously selected S. cerevisiae populations under similar selection, despite their >100-million-year divergence. We find K. lactis multicellularity is conferred by mutations in genes ACE2 or AIM44, with ACE2 being predominant. They are a subset of the six genes involved in the S. cerevisiae multicellularity. Both ACE2 and AIM44 regulate cell division, indicating that the genetic convergence is likely due to conserved cellular replication mechanisms. Complex population dynamics involving multiple ACE2/AIM44 genotypes are found in most K. lactis lineages. The results show common ancestry and natural selection shape convergence while chance and contingency determine the degree of divergence.

Changing fitness effects of mutations through long-term bacterial evolution

The distribution of fitness effects of new mutations shapes evolution, but it is challenging to observe how it changes as organisms adapt. Using Escherichia coli lineages spanning 50,000 generations of evolution, we quantify the fitness effects of insertion mutations in every gene. Macroscopically, the fraction of deleterious mutations changed little over time whereas the beneficial tail declined sharply, approaching an exponential distribution. Microscopically, changes in individual gene essentiality and deleterious effects often occurred in parallel; altered essentiality is only partly explained by structural variation. The identity and effect sizes of beneficial mutations changed rapidly over time, but many targets of selection remained predictable because of the importance of loss-of-function mutations. Taken together, these results reveal the dynamic-but statistically predictable-nature of mutational fitness effects.

Mutation bias and adaptation in bacteria

Genetic mutation, which provides the raw material for evolutionary adaptation, is largely a stochastic force. However, there is ample evidence showing that mutations can also exhibit strong biases, with some mutation types and certain genomic positions mutating more often than others. It is becoming increasingly clear that mutational bias can play a role in determining adaptive outcomes in bacteria in both the laboratory and the clinic. As such, understanding the causes and consequences of mutation bias can help microbiologists to anticipate and predict adaptive outcomes. In this review, we provide an overview of the mechanisms and features of the bacterial genome that cause mutational biases to occur. We then describe the environmental triggers that drive these mechanisms to be more potent and outline the adaptive scenarios where mutation bias can synergize with natural selection to define evolutionary outcomes. We conclude by describing how understanding mutagenic genomic features can help microbiologists predict areas sensitive to mutational bias, and finish by outlining future work that will help us achieve more accurate evolutionary forecasts.

Mutational hotspots lead to robust but suboptimal adaptive outcomes in certain environments

The observed mutational spectrum of adaptive outcomes can be constrained by many factors. For example, mutational biases can narrow the observed spectrum by increasing the rate of mutation at isolated sites in the genome. In contrast, complex environments can shift the observed spectrum by defining fitness consequences of mutational routes. We investigate the impact of different nutrient environments on the evolution of motility in Pseudomonas fluorescens Pf0-2x (an engineered non-motile derivative of Pf0-1) in the presence and absence of a strong mutational hotspot. Previous work has shown that this mutational hotspot can be built and broken via six silent mutations, which provide rapid access to a mutation that rescues swimming motility and confers the strongest swimming phenotype in specific environments. Here, we evolved a hotspot and non-hotspot variant strain of Pf0-2x for motility under nutrient-rich (LB) and nutrient-limiting (M9) environmental conditions. We observed the hotspot strain consistently evolved faster across all environmental conditions and its mutational spectrum was robust to environmental differences. However, the non-hotspot strain had a distinct mutational spectrum that changed depending on the nutrient environment. Interestingly, while alternative adaptive mutations in nutrient-rich environments were equal to, or less effective than, the hotspot mutation, the majority of these mutations in nutrient-limited conditions produced superior swimmers. Our competition experiments mirrored these findings, underscoring the role of environment in defining both the mutational spectrum and the associated phenotype strength. This indicates that while mutational hotspots working in concert with natural selection can speed up access to robust adaptive mutations (which can provide a competitive advantage in evolving populations), they can limit exploration of the mutational landscape, restricting access to potentially stronger phenotypes in specific environments.

The influence of immune cells on the existence of virus quasi-species

This article investigate a nonlocal reaction-diffusion system of equations modeling virus distribution with respect to their genotypes in the interaction with the immune response. This study demonstrates the existence of pulse solutions corresponding to virus quasi-species. The proof is based on the Leray-Schauder method, which relies on the topological degree for elliptic operators in unbounded domains and a priori estimates of solutions. Furthermore, linear stability analysis of a spatially homogeneous stationary solution identifies the critical conditions for the emergence of spatial and spatiotemporal structures. Finally, numerical simulations are used to illustrate nonlinear dynamics and pattern formation in the nonlocal model.

Predicting evolutionary outcomes through the probability of accessing sequence variants

Natural selection can only operate on available genetic variation. Thus, determining the probability of accessing different sequence variants from a starting sequence can help predict evolutionary trajectories and outcomes. We define the concept of "variant accessibility" as the probability that a set of genotypes encoding a particular protein function will arise through mutations before subject to natural selection. This probability is shaped by the mutational biases of nucleotides and the structure of the genetic code. Using the influenza A virus as a model, we discuss how a more accessible but less fit variant can emerge as an adaptation rather than a more fit variant. We describe a genotype-accessibility landscape, complementary to the genotype-fitness landscape, that informs the likelihood of a starting sequence reaching different parts of genotype space. The proposed framework lays the foundation for predicting the emergence of adaptive genotypes in evolving systems such as viruses and tumors.

Evolution of cyclic di-GMP signalling on a short and long term time scale

Diversifying radiation of domain families within specific lineages of life indicates the importance of their functionality for the organisms. The foundation for the diversifying radiation of the cyclic di-GMP signalling network that occurred within the bacterial kingdom is most likely based in the outmost adaptability, flexibility and plasticity of the system. Integrative sensing of multiple diverse extra- and intracellular signals is made possible by the N-terminal sensory domains of the modular cyclic di-GMP turnover proteins, mutations in the protein scaffolds and subsequent signal reception by diverse receptors, which eventually rewires opposite host-associated as well as environmental life styles including parallel regulated target outputs. Natural, laboratory and microcosm derived microbial variants often with an altered multicellular biofilm behaviour as reading output demonstrated single amino acid substitutions to substantially alter catalytic activity including substrate specificity. Truncations and domain swapping of cyclic di-GMP signalling genes and horizontal gene transfer suggest rewiring of the network. Presence of cyclic di-GMP signalling genes on horizontally transferable elements in particular observed in extreme acidophilic bacteria indicates that cyclic di-GMP signalling and biofilm components are under selective pressure in these types of environments. On a short and long term evolutionary scale, within a species and in families within bacterial orders, respectively, the cyclic di-GMP signalling network can also rapidly disappear. To investigate variability of the cyclic di-GMP signalling system on various levels will give clues about evolutionary forces and discover novel physiological and metabolic pathways affected by this intriguing second messenger signalling system.

Transient mutation bias increases the predictability of evolution on an empirical genotype-phenotype landscape

Predicting how a population will likely navigate a genotype-phenotype landscape requires consideration of selection in combination with mutation bias, which can skew the likelihood of following a particular trajectory. Strong and persistent directional selection can drive populations to ascend toward a peak. However, with a greater number of peaks and more routes to reach them, adaptation inevitably becomes less predictable. Transient mutation bias, which operates only on one mutational step, can influence landscape navigability by biasing the mutational trajectory early in the adaptive walk. This sets an evolving population upon a particular path, constraining the number of accessible routes and making certain peaks and routes more likely to be realized than others. In this work, we employ a model system to investigate whether such transient mutation bias can reliably and predictably place populations on a mutational trajectory to the strongest selective phenotype or usher populations to realize inferior phenotypic outcomes. For this we use motile mutants evolved from ancestrally non-motile variants of the microbe Pseudomonas fluorescens SBW25, of which one trajectory exhibits significant mutation bias. Using this system, we elucidate an empirical genotype-phenotype landscape, where the hill-climbing process represents increasing strength of the motility phenotype, to reveal that transient mutation bias can facilitate rapid and predictable ascension to the strongest observed phenotype in place of equivalent and inferior trajectories. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

Mutation bias and the predictability of evolution

Predicting evolutionary outcomes is an important research goal in a diversity of contexts. The focus of evolutionary forecasting is usually on adaptive processes, and efforts to improve prediction typically focus on selection. However, adaptive processes often rely on new mutations, which can be strongly influenced by predictable biases in mutation. Here, we provide an overview of existing theory and evidence for such mutation-biased adaptation and consider the implications of these results for the problem of prediction, in regard to topics such as the evolution of infectious diseases, resistance to biochemical agents, as well as cancer and other kinds of somatic evolution. We argue that empirical knowledge of mutational biases is likely to improve in the near future, and that this knowledge is readily applicable to the challenges of short-term prediction. This article is part of the theme issue 'Interdisciplinary approaches to predicting evolutionary biology'.

Distribution of mutation rates challenges evolutionary predictability

Natural selection is commonly assumed to act on extensive standing genetic variation. Yet, accumulating evidence highlights the role of mutational processes creating this genetic variation: to become evolutionarily successful, adaptive mutants must not only reach fixation, but also emerge in the first place, i.e. have a high enough mutation rate. Here, we use numerical simulations to investigate how mutational biases impact our ability to observe rare mutational pathways in the laboratory and to predict outcomes in experimental evolution. We show that unevenness in the rates at which mutational pathways produce adaptive mutants means that most experimental studies lack power to directly observe the full range of adaptive mutations. Modelling mutation rates as a distribution, we show that a substantially larger target size ensures that a pathway mutates more commonly. Therefore, we predict that commonly mutated pathways are conserved between closely related species, but not rarely mutated pathways. This approach formalizes our proposal that most mutations have a lower mutation rate than the average mutation rate measured experimentally. We suggest that the extent of genetic variation is overestimated when based on the average mutation rate.

Integrating pH into the metabolic theory of ecology to predict bacterial diversity in soil

Microorganisms play essential roles in soil ecosystem functioning and maintenance, but methods are currently lacking for quantitative assessments of the mechanisms underlying microbial diversity patterns observed across disparate systems and scales. Here we established a quantitative model to incorporate pH into metabolic theory to capture and explain some of the unexplained variation in the relationship between temperature and soil bacterial diversity. We then tested and validated our newly developed models across multiple scales of ecological organization. At the species level, we modeled the diversification rate of the model bacterium Pseudomonas fluorescens evolving under laboratory media gradients varying in temperature and pH. At the community level, we modeled patterns of bacterial communities in paddy soils across a continental scale, which included natural gradients of pH and temperature. Last, we further extended our model at a global scale by integrating a meta-analysis comprising 870 soils collected worldwide from a wide range of ecosystems. Our results were robust in consistently predicting the distributional patterns of bacterial diversity across soil temperature and pH gradients-with model variation explaining from 7 to 66% of the variation in bacterial diversity, depending on the scale and system complexity. Together, our study represents a nexus point for the integration of soil bacterial diversity and quantitative models with the potential to be used at distinct spatiotemporal scales. By mechanistically representing pH into metabolic theory, our study enhances our capacity to explain and predict the patterns of bacterial diversity and functioning under current or future climate change scenarios.

Towards evolutionary predictions: Current promises and challenges

Evolution has traditionally been a historical and descriptive science, and predicting future evolutionary processes has long been considered impossible. However, evolutionary predictions are increasingly being developed and used in medicine, agriculture, biotechnology and conservation biology. Evolutionary predictions may be used for different purposes, such as to prepare for the future, to try and change the course of evolution or to determine how well we understand evolutionary processes. Similarly, the exact aspect of the evolved population that we want to predict may also differ. For example, we could try to predict which genotype will dominate, the fitness of the population or the extinction probability of a population. In addition, there are many uses of evolutionary predictions that may not always be recognized as such. The main goal of this review is to increase awareness of methods and data in different research fields by showing the breadth of situations in which evolutionary predictions are made. We describe how diverse evolutionary predictions share a common structure described by the predictive scope, time scale and precision. Then, by using examples ranging from SARS-CoV2 and influenza to CRISPR-based gene drives and sustainable product formation in biotechnology, we discuss the methods for predicting evolution, the factors that affect predictability and how predictions can be used to prevent evolution in undesirable directions or to promote beneficial evolution (i.e. evolutionary control). We hope that this review will stimulate collaboration between fields by establishing a common language for evolutionary predictions.

The rapid evolution of flagellar ion selectivity in experimental populations of E. coli

Determining which cellular processes facilitate adaptation requires a tractable experimental model where an environmental cue can generate variants that rescue function. The bacterial flagellar motor (BFM) is an excellent candidate-an ancient and highly conserved molecular complex for bacterial propulsion toward favorable environments. Motor rotation is often powered by H⁺ or Na⁺ ion transit through the torque-generating stator subunit of the motor complex, and ion selectivity has adapted over evolutionary time scales. Here, we used CRISPR engineering to replace the native Escherichia coli H⁺-powered stator with Na⁺-powered stator genes and report the spontaneous reversion of our edit in a low-sodium environment. We followed the evolution of the stators during their reversion to H⁺-powered motility and used both whole-genome and RNA sequencing to identify genes involved in the cell's adaptation. Our transplant of an unfit protein and the cells' rapid response to this edit demonstrate the adaptability of the stator subunit and highlight the hierarchical modularity of the flagellar motor.

Population size mediates the contribution of high-rate and large-benefit mutations to parallel evolution

Mutations with large fitness benefits and mutations occurring at high rates may both cause parallel evolution, but their contribution is predicted to depend on population size. Moreover, high-rate and large-benefit mutations may have different long-term adaptive consequences. We show that small and 100-fold larger bacterial populations evolve resistance to a β-lactam antibiotic by using similar numbers, but different types of mutations. Small populations frequently substitute similar high-rate structural variants and loss-of-function point mutations, including the deletion of a low-activity β-lactamase, and evolve modest resistance levels. Large populations more often use low-rate, large-benefit point mutations affecting the same targets, including mutations activating the β-lactamase and other gain-of-function mutations, leading to much higher resistance levels. Our results demonstrate the separation by clonal interference of mutation classes with divergent adaptive consequences, causing a shift from high-rate to large-benefit mutations with increases in population size.

Evolutionary flexibility in routes to mat formation by Pseudomonas

Many bacteria form mats at the air-liquid interface of static microcosms. These structures typically involve the secretion of exopolysaccharides, the production of which is often controlled by the secondary messenger c-di-GMP. Mechanisms of mat formation have been particularly well characterized in Pseudomonas fluorescens SBW25; stimuli or mutations that increase c-di-GMP production by diguanylate cyclases (WspR, AwsR, and MwsR) result in the secretion of cellulose and mat formation. Here, we characterize and compare mat formation in two close relatives of SBW25: Pseudomonas simiae PICF7 and P. fluorescens A506. We find that PICF7-the strain more closely related to SBW25-can form mats through mutations affecting the activity of the same three diguanylate cyclases as SBW25. However, instead of cellulose, these mutations activate production of the exopolysaccharide Pel. We also provide evidence for at least two further-as yet uncharacterized-routes to mat formation by PICF7. P. fluorescens A506, while retaining the same mutational routes to mat formation as SBW25 and PICF7, preferentially forms mats by a semi-heritable mechanism that culminates in Psl and Pga over-production. Our results demonstrate a high level of evolutionary flexibility in the molecular and structural routes to mat formation, even among close relatives.

Rugose small colony variant and its hyper-biofilm in Pseudomonas aeruginosa: Adaption, evolution, and biotechnological potential

One of the hallmarks of the environmental bacterium Pseudomonas aeruginosa is its excellent ecological flexibility, which can thrive in diverse ecological niches. In different ecosystems, P. aeruginosa may use different strategies to survive, such as forming biofilms in crude oil environment, converting to mucoid phenotype in the cystic fibrosis (CF) lung, or becoming persisters when treated with antibiotics. Rugose small colony variants (RSCVs) are the adaptive mutants of P. aeruginosa, which can be frequently isolated from chronic infections. During the past years, there has been a renewed interest in using P. aeruginosa as a model organism to investigate the RSCVs formation, persistence and pathogenesis, as RSCVs represent a hyper-biofilm formation, high adaptability, high-tolerance sub-population in biofilms. This review will briefly summarize recent advances regarding the phenotypic, genetic and host interaction associated with RSCVs, with an emphasis on P. aeruginosa. Meanwhile, some non-pathogenic bacteria such as Pseudomonas fluorescence, Pseudomonas putida and Bacillus subtilis will be also included. Remarkable emphasis is given on intrinsic functions of such hyper-biofilm formation characteristic as well as its potential applications in several biocatalytic transformations including wastewater treatment, microbial fermentation, and plastic degradation. Hopefully, this review will attract the interest of researchers in various fields and shape future research focused not only on evolutionary biology but also on biotechnological applications related to RSCVs.

A mutational hotspot that determines highly repeatable evolution can be built and broken by silent genetic changes

Mutational hotspots can determine evolutionary outcomes and make evolution repeatable. Hotspots are products of multiple evolutionary forces including mutation rate heterogeneity, but this variable is often hard to identify. In this work, we reveal that a near-deterministic genetic hotspot can be built and broken by a handful of silent mutations. We observe this when studying homologous immotile variants of the bacteria Pseudomonas fluorescens, AR2 and Pf0-2x. AR2 resurrects motility through highly repeatable de novo mutation of the same nucleotide in >95% lines in minimal media (ntrB A289C). Pf0-2x, however, evolves via a number of mutations meaning the two strains diverge significantly during adaptation. We determine that this evolutionary disparity is owed to just 6 synonymous variations within the ntrB locus, which we demonstrate by swapping the sites and observing that we are able to both break (>95% to 0%) and build (0% to 80%) a deterministic mutational hotspot. Our work reveals a key role for silent genetic variation in determining adaptive outcomes.

Mutations in surface-sensing receptor WspA lock the Wsp signal transduction system into a constitutively active state

Pseudomonas aeruginosa rugose small-colony variants (RSCVs) are frequently isolated from chronic infections, yet, they are rarely reported in environmental isolates. Here, during the comparative genomic analysis of two P. aeruginosa strains isolated from crude oil, we discovered a spontaneous in-frame deletion, wspA_Δ280-307 , which led to hyper-biofilm and RSCV phenotypes. WspA is a homologue of methyl-accepting chemotaxis proteins (MCPs) that senses surfaces to regulate biofilm formation by stimulating cyclic-di-guanosine monophosphate (c-di-GMP) synthesis through the Wsp system. However, the methylation sites of WspA have never been identified. In this study, we identified E280 and E294 of WspA as methylation sites. The wspA_Δ280-307 mutation enabled the Wsp system to lock into a constitutively active state that is independent of regulation by methylation. The result is an enhanced production of c-di-GMP. Sequence alignment revealed three conserved repeat sequences within the amino acid residues 280-313 (aa280-313) region of WspA homologues, suggesting that a spontaneous deletion within this DNA encoding region was likely a result of intragenic recombination and that similar mutations might occur in several related bacterial genera. Our results provide a plausible explanation for the selection of RSCVs and a mechanism to confer a competitive advantage for P. aeruginosa in a crude-oil environment.

Can gene-inactivating mutations lead to evolutionary novelty?

Evolutionary novelty is difficult to define. It typically involves shifts in organismal or biochemical phenotypes that can be seen as qualitative as well as quantitative changes. In laboratory-based experimental evolution of novel phenotypes and the human domestication of crops, the majority of the mutations that lead to adaptation are loss-of-function mutations that impair or eliminate the function of genes rather than gain-of-function mutations that increase or qualitatively alter the function of proteins. Here, I speculate that easier access to loss-of-function mutations has led them to play a major role in the adaptive radiations that occur when populations have access to many unoccupied ecological niches. I discuss five possible objections to this claim: that genes can only survive if they confer benefits to the organisms that bear them, antagonistic pleiotropy, the importance of pre-existing genetic variation in populations, the danger that adaptation by breaking genes will, over long times, cause organisms to run out of genes, and the recessive nature of most loss-of-function mutations.

Forecasting of phenotypic and genetic outcomes of experimental evolution in Pseudomonas protegens

Experimental evolution with microbes is often highly repeatable under identical conditions, suggesting the possibility to predict short-term evolution. However, it is not clear to what degree evolutionary forecasts can be extended to related species in non-identical environments, which would allow testing of general predictive models and fundamental biological assumptions. To develop an extended model system for evolutionary forecasting, we used previous data and models of the genotype-to-phenotype map from the wrinkly spreader system in Pseudomonas fluorescens SBW25 to make predictions of evolutionary outcomes on different biological levels for Pseudomonas protegens Pf-5. In addition to sequence divergence (78% amino acid and 81% nucleotide identity) for the genes targeted by mutations, these species also differ in the inability of Pf-5 to make cellulose, which is the main structural basis for the adaptive phenotype in SBW25. The experimental conditions were changed compared to the SBW25 system to test if forecasts were extendable to a non-identical environment. Forty-three mutants with increased ability to colonize the air-liquid interface were isolated, and the majority had reduced motility and was partly dependent on the Pel exopolysaccharide as a structural component. Most (38/43) mutations are expected to disrupt negative regulation of the same three diguanylate cyclases as in SBW25, with a smaller number of mutations in promoter regions, including an uncharacterized polysaccharide synthase operon. A mathematical model developed for SBW25 predicted the order of the three main pathways and the genes targeted by mutations, but differences in fitness between mutants and mutational biases also appear to influence outcomes. Mutated regions in proteins could be predicted in most cases (16/22), but parallelism at the nucleotide level was low and mutational hot spot sites were not conserved. This study demonstrates the potential of short-term evolutionary forecasting in experimental populations and provides testable predictions for evolutionary outcomes in other Pseudomonas species.

Eco-evolutionary interaction between microbiome presence and rapid biofilm evolution determines plant host fitness

Microbiomes are important to the survival and reproduction of their hosts. Although ecological and evolutionary processes can happen simultaneously in microbiomes, little is known about how microbiome eco-evolutionary dynamics determine host fitness. Here we show, using experimental evolution, that fitness of the aquatic plant Lemna minor is modified by interactions between the microbiome and the evolution of one member, Pseudomonas fluorescens. Microbiome presence promotes P. fluorescens' rapid evolution to form biofilm, which reciprocally alters the microbiome's species composition. These eco-evolutionary dynamics modify the host's multigenerational fitness. The microbiome and non-evolving P. fluorescens together promote host fitness, whereas the microbiome with P. fluorescens that evolves biofilm reduces the beneficial impact on host fitness. Additional experiments suggest that the microbial effect on host fitness may occur through changes in microbiome production of auxin, a plant growth hormone. Our study, therefore, experimentally demonstrates the importance of the eco-evolutionary dynamics in microbiomes for host-microbiome interactions.

Dynamics of bacterial adaptation

Determining pattern in the dynamics of population evolution is a long-standing focus of evolutionary biology. Complementing the study of natural populations, microbial laboratory evolution experiments have become an important tool for addressing these dynamics because they allow detailed and replicated analysis of evolution in response to controlled environmental and genetic conditions. Key findings include a tendency for smoothly declining rates of adaptation during selection in constant environments, at least in part a reflection of antagonism between accumulating beneficial mutations, and a large number of beneficial mutations available to replicate populations leading to significant, but relatively low genetic parallelism, even as phenotypic characteristics show high similarity. Together, there is a picture of adaptation as a process with a varied and largely unpredictable genetic basis leading to much more similar phenotypic outcomes. Increasing sophistication of sequencing and genetic tools will allow insight into mechanisms behind these and other patterns.

Multi-scale Predictions of Drug Resistance Epidemiology Identify Design Principles for Rational Drug Design

Rationally designing drugs that last longer in the face of biological evolution is a critical objective of drug discovery. However, this goal is thwarted by the diversity and stochasticity of evolutionary trajectories that drive uncertainty in the clinic. Although biophysical models can qualitatively predict whether a mutation causes resistance, they cannot quantitatively predict the relative abundance of resistance mutations in patient populations. We present stochastic, first-principle models that are parameterized on a large in vitro dataset and that accurately predict the epidemiological abundance of resistance mutations across multiple leukemia clinical trials. The ability to forecast resistance variants requires an understanding of their underlying mutation biases. Beyond leukemia, a meta-analysis across prostate cancer, breast cancer, and gastrointestinal stromal tumors suggests that resistance evolution in the adjuvant setting is influenced by mutational bias. Our analysis establishes a principle for rational drug design: when evolution favors the most probable mutant, so should drug design.

Drug resistance is often addressed through next-generation drug design, but evolutionary diversity complicates these efforts. Here, Leighow et al. demonstrate that multi-scale models can quantitatively predict mutant frequency. We find that when heterogeneity is limited, analysis requires an understanding of substitution likelihood. We show that these models can inform evolutionarily optimized drug design.

Inferring a complete genotype-phenotype map from a small number of measured phenotypes

Understanding evolution requires detailed knowledge of genotype-phenotype maps; however, it can be a herculean task to measure every phenotype in a combinatorial map. We have developed a computational strategy to predict the missing phenotypes from an incomplete, combinatorial genotype-phenotype map. As a test case, we used an incomplete genotype-phenotype dataset previously generated for the malaria parasite’s ‘chloroquine resistance transporter’ (PfCRT). Wild-type PfCRT (PfCRT^3D7) lacks significant chloroquine (CQ) transport activity, but the introduction of the eight mutations present in the ‘Dd2’ isoform of PfCRT (PfCRT^Dd2) enables the protein to transport CQ away from its site of antimalarial action. This gain of a transport function imparts CQ resistance to the parasite. A combinatorial map between PfCRT^3D7 and PfCRT^Dd2 consists of 256 genotypes, of which only 52 have had their CQ transport activities measured through expression in the Xenopus laevis oocyte. We trained a statistical model with these 52 measurements to infer the CQ transport activity for the remaining 204 combinatorial genotypes between PfCRT^3D7 and PfCRT^Dd2. Our best-performing model incorporated a binary classifier, a nonlinear scale, and additive effects for each mutation. The addition of specific pairwise- and high-order-epistatic coefficients decreased the predictive power of the model. We evaluated our predictions by experimentally measuring the CQ transport activities of 24 additional PfCRT genotypes. The R² value between our predicted and newly-measured phenotypes was 0.90. We then used the model to probe the accessibility of evolutionary trajectories through the map. Approximately 1% of the possible trajectories between PfCRT^3D7 and PfCRT^Dd2 are accessible; however, none of the trajectories entailed eight successive increases in CQ transport activity. These results demonstrate that phenotypes can be inferred with known uncertainty from a partial genotype-phenotype dataset. We also validated our approach against a collection of previously published genotype-phenotype maps. The model therefore appears general and should be applicable to a large number of genotype-phenotype maps.

Biological macromolecules are built from chains of building blocks. The function of a macromolecule depends on the specific chemical properties of the building blocks that make it up. Macromolecules evolve through mutations that swap one building block for another. Understanding how biomolecules work and evolve therefore requires knowledge of the effects of mutations. The effects of mutations can be measured experimentally; however, because there are a vast number of possible combinations of mutations, it is often difficult to make enough measurements to understand biomolecular function and evolution. In this paper, we describe a simple method to predict the effects of mutations on biomolecules from a small number of measurements. This method works by appropriately averaging the effects of mutations seen in different contexts. We test the method by predicting the effects of mutations on a PfCRT—a macromolecule from the malarial parasite that confers drug resistance. We find that our method is fast and effective. Using a small number of measurements, we were able to gain insight into the evolutionary steps by which this macromolecule conferred drug resistance. To make this method accessible to other researchers, we have released it as an open-source software package: https://gpseer.readthedocs.io.

How Do Cells Adapt? Stories Told in Landscapes

Cells adapt to changing environments. Perturb a cell and it returns to a point of homeostasis. Perturb a population and it evolves toward a fitness peak. We review quantitative models of the forces of adaptation and their visualizations on landscapes. While some adaptations result from single mutations or few-gene effects, others are more cooperative, more delocalized in the genome, and more universal and physical. For example, homeostasis and evolution depend on protein folding and aggregation, energy and protein production, protein diffusion, molecular motor speeds and efficiencies, and protein expression levels. Models provide a way to learn about the fitness of cells and cell populations by making and testing hypotheses.

Hypermutator Pseudomonas aeruginosa Exploits Multiple Genetic Pathways To Develop Multidrug Resistance during Long-Term Infections in the Airways of Cystic Fibrosis Patients

Pseudomonas aeruginosa exploits intrinsic and acquired resistance mechanisms to resist almost every antibiotic used in chemotherapy. Antimicrobial resistance in P. aeruginosa isolates recovered from cystic fibrosis (CF) patients is further enhanced by the occurrence of hypermutator strains, a hallmark of chronic infections in CF patients. However, the within-patient genetic diversity of P. aeruginosa populations related to antibiotic resistance remains unexplored. Here, we show the evolution of the mutational resistome profile of a P. aeruginosa hypermutator lineage by performing longitudinal and transversal analyses of isolates collected from a CF patient throughout 20 years of chronic infection. Our results show the accumulation of thousands of mutations, with an overall evolutionary history characterized by purifying selection. However, mutations in antibiotic resistance genes appear to have been positively selected, driven by antibiotic treatment. Antibiotic resistance increased as infection progressed toward the establishment of a population constituted by genotypically diversified coexisting sublineages, all of which converged to multidrug resistance. These sublineages emerged by parallel evolution through distinct evolutionary pathways, which affected genes of the same functional categories. Interestingly, ampC and ftsI, encoding the β-lactamase and penicillin-binding protein 3, respectively, were found to be among the most frequently mutated genes. In fact, both genes were targeted by multiple independent mutational events, which led to a wide diversity of coexisting alleles underlying β-lactam resistance. Our findings indicate that hypermutators, apart from boosting antibiotic resistance evolution by simultaneously targeting several genes, favor the emergence of adaptive innovative alleles by clustering beneficial/compensatory mutations in the same gene, hence expanding P. aeruginosa strategies for persistence.

Emerging Frontiers in the Study of Molecular Evolution

A collection of the editors of Journal of Molecular Evolution have gotten together to pose a set of key challenges and future directions for the field of molecular evolution. Topics include challenges and new directions in prebiotic chemistry and the RNA world, reconstruction of early cellular genomes and proteins, macromolecular and functional evolution, evolutionary cell biology, genome evolution, molecular evolutionary ecology, viral phylodynamics, theoretical population genomics, somatic cell molecular evolution, and directed evolution. While our effort is not meant to be exhaustive, it reflects research questions and problems in the field of molecular evolution that are exciting to our editors.

Probabilistic Models for Predicting Mutational Routes to New Adaptive Phenotypes

Understanding the translation of genetic variation to phenotypic variation is a fundamental problem in genetics and evolutionary biology. The introduction of new genetic variation through mutation can lead to new adaptive phenotypes, but the complexity of the genotype-to-phenotype map makes it challenging to predict the phenotypic effects of mutation. Metabolic models, in conjunction with flux balance analysis, have been used to predict evolutionary optimality. These methods however rely on large scale models of metabolism, describe a limited set of phenotypes, and assume that selection for growth rate is the prime evolutionary driver. Here we describe a method for computing the relative likelihood that mutational change will translate into a phenotypic change between two molecular pathways. The interactions of molecular components in the pathways are modeled with ordinary differential equations. Unknown parameters are offset by probability distributions that describe the concentrations of molecular components, the reaction rates for different molecular processes, and the effects of mutations. Finally, the likelihood that mutations in a pathway will yield phenotypic change is estimated with stochastic simulations. One advantage of this method is that only basic knowledge of the interaction network underlying a phenotype is required. However, it can also incorporate available information about concentrations and reaction rates as well as mutational biases and mutational robustness of molecular components. The method estimates the relative probabilities that different pathways produce phenotypic change, which can be combined with fitness models to predict evolutionary outcomes.

Causes and Biophysical Consequences of Cellulose Production by Pseudomonas fluorescens SBW25 at the Air-Liquid Interface

This work reveals a hitherto unrecognized behavior that manifests at the air-liquid interface that depends on production of cellulose and hints at undiscovered dimensions to bacterial life at surfaces. Additionally, the study links activation of known diguanylate cyclase-encoding pathways to cellulose expression and to signals encountered at the meniscus. Further significance stems from recognition of the consequences of fluid instabilities arising from surface production of cellulose for transport of water-soluble products over large distances.

Cellulose-overproducing wrinkly spreader mutants of Pseudomonas fluorescens SBW25 have been the focus of much investigation, but conditions promoting the production of cellulose in ancestral strain SBW25 and its effects and consequences have escaped in-depth investigation through lack of an in vitro phenotype. Here, using a custom-built device, we reveal that in static broth microcosms, ancestral SBW25 encounters environmental signals at the air-liquid interface that activate, via three diguanylate cyclase-encoding pathways (Wsp, Aws, and Mws), production of cellulose. Secretion of the polymer at the meniscus leads to modification of the environment and growth of numerous microcolonies that extend from the surface. Accumulation of cellulose and associated microbial growth leads to Rayleigh-Taylor instability resulting in bioconvection and rapid transport of water-soluble products over tens of millimeters. Drawing upon data, we built a mathematical model that recapitulates experimental results and captures the interactions between biological, chemical and physical processes.

IMPORTANCE This work reveals a hitherto unrecognized behavior that manifests at the air-liquid interface that depends on production of cellulose and hints at undiscovered dimensions to bacterial life at surfaces. Additionally, the study links activation of known diguanylate cyclase-encoding pathways to cellulose expression and to signals encountered at the meniscus. Further significance stems from recognition of the consequences of fluid instabilities arising from surface production of cellulose for transport of water-soluble products over large distances.

Pseudomonas aeruginosa Interstrain Dynamics and Selection of Hyperbiofilm Mutants during a Chronic Infection

Opportunistic pathogens establishing new infections experience strong selection to adapt, often favoring mutants that persist. Capturing this initial dynamic is critical for identifying the first adaptations that drive pathogenesis. Here we used a porcine full-thickness burn wound model of chronic infection to study the evolutionary dynamics of diverse Pseudomonas aeruginosa infections. Wounds were infected with a mixed community of six P. aeruginosa strains, including the model PA14 strain (PA14-1), and biopsies taken at 3, 14, and 28 days postinfection. Hyperbiofilm-forming rugose small-colony variants (RSCVs) were the earliest and predominant phenotypic variant. These variants were detected on day 3 and persisted, with the majority evolved from PA14-1. Whole-genome sequencing of PA14-1 RSCV isolates revealed driver mutations exclusively in the wsp pathway, conferring hyperbiofilm phenotypes. Several of the wsp mutant RSCVs also acquired CRISPR-Cas adaptive immunity to prophages isolated from the P. aeruginosa wound isolate (B23-2) that was also present in the inoculum. These observations emphasize the importance of interstrain dynamics and the role of lysogenic phages in the survival of an invading pathogen. Rather than being a side effect of chronicity, the rapid rise of RSCVs in wounds is evidence of positive selection on the Wsp chemosensory system to produce mutants with elevated biofilm formation capacity. We predict that RSCVs provide a level of phenotypic diversity to the infecting bacterial community and are common, early adaptations during infections. This would likely have significant consequences for clinical outcomes.IMPORTANCE Bacteria adapt to infections by evolving variants that are more fit and persistent. These recalcitrant variants are typically observed in chronic infections. However, it is unclear when and why these variants evolve. To address these questions, we used a porcine chronic wound model to study the evolutionary dynamics of Pseudomonas aeruginosa in a mixed-strain infection. We isolated hyperbiofilm variants that persisted early in the infection. Interstrain interactions were also observed, where adapted variants acquired CRISPR-mediated immunity to phages. We show that when initiating infection, P. aeruginosa experiences strong positive selection for hyperbiofilm phenotypes produced by mutants of a single chemosensory system, the Wsp pathway. We predict that hyperbiofilm variants are early adaptations to infection and that interstrain interactions may influence bacterial burden and infection outcomes.

Repeated Phenotypic Evolution by Different Genetic Routes in Pseudomonas fluorescens SBW25

Repeated evolution of functionally similar phenotypes is observed throughout the tree of life. The extent to which the underlying genetics are conserved remains an area of considerable interest. Previously, we reported the evolution of colony switching in two independent lineages of Pseudomonas fluorescens SBW25. The phenotypic and genotypic bases of colony switching in the first lineage (Line 1) have been described elsewhere. Here, we deconstruct the evolution of colony switching in the second lineage (Line 6). We show that, as for Line 1, Line 6 colony switching results from an increase in the expression of a colanic acid-like polymer (CAP). At the genetic level, nine mutations occur in Line 6. Only one of these—a nonsynonymous point mutation in the housekeeping sigma factor rpoD—is required for colony switching. In contrast, the genetic basis of colony switching in Line 1 is a mutation in the metabolic gene carB. A molecular model has recently been proposed whereby the carB mutation increases capsulation by redressing the intracellular balance of positive (ribosomes) and negative (RsmAE/CsrA) regulators of a positive feedback loop in capsule expression. We show that Line 6 colony switching is consistent with this model; the rpoD mutation generates an increase in ribosomal gene expression, and ultimately an increase in CAP expression.

Parallel evolution of HIV-1 in a long-term experiment

One of the most intriguing puzzles in biology is the degree to which evolution is repeatable. The repeatability of evolution, or parallel evolution, has been studied in a variety of model systems, but has rarely been investigated with clinically relevant viruses. To investigate parallel evolution of HIV-1, we passaged two replicate HIV-1 populations for almost 1 year in each of two human T-cell lines. For each of the four evolution lines, we determined the genetic composition of the viral population at nine time points by deep sequencing the entire genome. Mutations that were carried by the majority of the viral population accumulated continuously over 1 year in each evolution line. Many majority mutations appeared in more than one evolution line, that is, our experiments showed an extreme degree of parallel evolution. In one of the evolution lines, 62% of the majority mutations also occur in another line. The parallelism impairs our ability to reconstruct the evolutionary history by phylogenetic methods. We show that one can infer the correct phylogenetic topology by including minority mutations in our analysis. We also find that mutation diversity at the beginning of the experiment is predictive of the frequency of majority mutations at the end of the experiment.