Previous two research highlights for Evolutionary Applications

For the past two research highlights at Evolutionary Applications, I first covered a great paper summarizing the many way evolutionary theory can be applied to current issues by Scot Carroll and colleagues:

“As we highlight each month in this section, the application of evolutionary theory to issues affecting the health and well-being of human, agricultural, and natural populations is gaining increasing momentum. In a recent review article written for Science, Scott Carroll et al. take on the now monumental task of synthesizing the many ways that evolutionary biology can be used to address global challenges (Carroll et al. 2014). They comprehensively explore the main problems being tackled with an evolutionary approach, ranging from populations evolving too quickly (such as emerging pathogens or pests evolving resistance to treatment) to populations not evolving quickly enough (for example those being negatively affected by human-mediated change).

The authors begin by identifying what they see as the two key paradigms of applied evolutionary biology: (i) managing contemporary evolution (i.e., manipulating the rapid evolutionary response of short-lived organisms with large population sizes, such as bacterial pathogens) and (ii) altering the phenotype–environment mismatch (i.e., responding to populations of long-lived organisms such as trees that are no longer well adapted to their local environment due to shifts in climatic conditions or changes in biotic interactions). As a great example of such a mismatch, the authors highlight the increasing rates of obesity, diabetes, and heart disease in the human populations as a result of a more sedentary lifestyle with diets rich in sugars and fat. They then identify a number of promising research avenues that either have addressed or have the potential to address current global challenges, covering a wide range of approaches including the use of genetic engineering to more appropriately match genomes to their environment, the use of ‘refuges’ in agriculture and combination treatments against pests and pathogens to hinder the evolution of resistance, and introducing nonlocal genotypes which are predicted to perform better under given environmental conditions into natural populations to increase local adaptation.

The article nicely separates these conceptual approaches into strategies for slowing unwanted evolution or directly influencing fitness of pests and pathogens, strategies for reducing the mismatch between phenotype and the local environment, and strategies for increasing group performance by selecting on group-level traits. For example, the authors discuss the success of artificially selecting for group yield in agricultural plots rather than individual fitness as a means for decreasing competition among plants. Critically, the piece also emphasizes the need to take a unified approach in meeting international objectives for sustainable development and suggests a need for stricter enforcement of guidelines in order to ensure best practice is achieved despite temptation to put profit or immediate success ahead of sustainable solutions.

Overall, the review acts as a unique and remarkable resource both for researchers and students who are new to the field of applied evolution and those who actively contribute to the field.

Carroll, S. P.P. S. JørgensenM. T. KinnisonC. T. BergstromR. F. DenisonP. GluckmanT. B. Smith et al. 2014Applying evolutionary biology to address global challengesScience 346:1245993.”

And then discussed recent applications in molecular evolution, including two new papers using comparative genomics of mosquitos to better understand the evolution of these important disease vectors:

“The study of changing sequence composition of DNA, RNA and proteins over time has offered some of the most fundamental insights into the evolutionary process to date. From understanding how populations and ultimately species diverge to the study of how particular selection pressures affect changes in genotype and phenotype, our knowledge of evolution would be a fraction of what it is now without the major advances made in the field of molecular evolution. Recent technological and bioinformatical improvements have continued to expand these insights, and have also offered key applications such as the ability to model and predict pathogen evolution, monitor the effective population size of threatened species, and help understand what constitutes a healthy microbiome.

Two recent studies, both led by Nora Besansky and published in Science, emphasize the power and challenges of comparative genomics when working to understand the evolution of disease vectors. First, Daniel Neafsey and colleagues report the sequencing, assembly, and comparison of genomes from 16 Anopheles mosquito species (Neafsey et al. 2014). As 11 of these species are considered major disease vectors, comparison among the genomes allowed the researchers to examine underlying genes that may be associated with vectoring capacity. The results suggest that, relative to the Drosophila genus, the Anopholes’ genomes are remarkably flexible, with rapid rates of gene loss/gain, increased loss of introns, and shuffling of genes on the X chromosome. The data suggest a mechanism for the observed functional diversity across the species, especially in those traits such as chemosensory ability that are associated with adaptation to host feeding and therefore disease vectoring. However, comparison among genomes was hampered by what are most likely high levels of interspecific gene flow, or introgression, as described in a separate paper by Michael Fontaine and coauthors (Fontaine et al. 2014). Depending on which genomic segment the authors used to build phylogenetic trees, a remarkably different pattern emerged; trees based on autosomal sequences tended to group the three major vectors of malaria together, while those built using the X chromosome suggest early radiation of these three species and persistent introgression on the autosomes. Together, these studies offer tantalizing hypotheses for the adaptive significance of among-species gene flow and genomic plasticity in allowing the Anopholes genus to act as vectors for a wide array of pathogens.

In addition to the increasing power of genomics and phylogenomics, the use of transcriptional profiling has also proven invaluable to the field. A recent review of novel insights gained through transcriptomic analyses of natural populations by Mariano Alvarez and collaborators highlights the utility of this approach in testing how genotype translates to phenotype, and how this translation is influenced by environment-specific gene expression (Alvarez et al. 2014). Such variation can have dramatic implications for the process of adaptation as well as our ability to predict the response of populations to rapid environmental changes such as those resulting from pathogens, pollutants, or climate change. More recent advancement in transcriptomics includes the ability to profile gene expression of single cells, as discussed by Nicola Crosetto and coauthors in a new paper reviewing recent progress in spatiotemporal transcriptomics (Crosetto et al. 2015). Among the many applications of this powerful approach to unravelling among-cell expression differences is the ability to examine heterogeneity of tumour cells to predict drug sensitivity of various cancers.

The use of sequence data to infer evolutionary processes is not limited to single species. Indeed, the use of metagenomics to infer the composition of species from environmental samples has greatly enhanced our understanding of microbial diversity. In its simplest form, metagenomic analysis allows for a culture-independent characterization of microbial community composition. This type of analysis has gained much recent attention for its application in understanding the microbiomes of eukaryotic species. For example, recent work by Julia Goodrich and colleagues examined how human genetics shapes the relative abundances of various gut bacteria by comparing microbiotas across 416 pairs of twins (Goodrich et al. 2014). The authors first discovered a clear heritability for a subset of bacterial taxa, most notably those from the family Christensenellaceae, which were also correlated with low host body-mass index (BMI). The authors then went a step further by adding a particular species of Christensenellaceae into an obese-associated microbiome and inoculating sterile mice with either the unaltered or altered microbial community. In this way, they were able to demonstrate not only correlation with host metabolism in humans but also to infer causation, as mice supplemented with this species showed reduced weight gain relative to those not receiving the supplement.

The simultaneous analysis of multiple genomes within a single environmental sample also allows for assessment of selection acting on genes shared by members of the community. A terrific example of this comes from recent work by Molly Gibson and collaborators who examined the so-called ‘resistome’ of microbial communities from soil and the human gut, in this case focusing on the genes conferring resistance against 18 antibiotics typically used in clinical settings (Gibson et al. 2015). The authors used a new database of protein families to assign antibiotic resistance functions to each metagenomic segment, and were able to demonstrate that the antibiotic resistance genes found in environmental versus human-associated microbiota were functionally different, perhaps suggesting less gene flow among these communities than previously thought.

Overall, the recent advancements in both omics and bioinformatics have been game-changing for the field of molecular evolution, and the application of such new approaches and technologies have only begun to surface. The potential for advancement in clinical and agricultural settings is already being realized, and application to the management of natural populations, including the spread of disease, is already following.

Alvarez, M.A. W. Schrey, and C. L. Richards2014Ten years of transcriptomics in wild populations: what have we learned about their ecology and evolution? Molecular Ecology, doi: 10.1111/mec.13055

Crosetto, N.M. Bienko, and A. van Oudenaarden2015Spatially resolved transcriptomics and beyondNature Reviews Genetics16:5766

Fontaine, M. C.J. B. PeaseA. SteeleR. M. WaterhouseD. E. NeafseyI. V. SharakhovX. Jiang et al. 2014Extensive introgression in a malaria vector species complex revealed by phylogenomicsScience 125852:4

Gibson, M. K.K. J. Forsberg, and G. Dantas2015Improved annotation of antibiotic resistance determinants reveals microbial resistomes cluster by ecologyThe ISME journal 9:207216

Goodrich, J. K.J. L. WatersA. C. PooleJ. L. SutterO. KorenR. BlekhmanM. Beaumont et al. 2014Human genetics shape the gut microbiomeCell 159:789799

Neafsey, D. E.R. M. WaterhouseM. R. AbaiS. S. AganezovM. A. AlekseyevJ. E. AllenJ. Amon et al. 2014Highly evolvable malaria vectors: the genomes of 16 Anopheles mosquitoesScience 347:1258522.”

The ever-evolving field of agriculture

For this month’s Evolutionary Applications research highlight, I discuss recent uses of evolutionary theory in driving agricultural practice:

“The earliest application of evolutionary theory, although unknowingly at the time, was artificial selection of crops and animals for food production. Ever increasing technical advances in breeding, genetic engineering and comparative genomics have since led to a rapid acceleration in the rate of such selection, although many of the basic principles underlying the process have remained the same over time. For example, whereas we used to inter-breed among genotypes and even species to generate standing genetic variation upon which to select, we can now introduce specific genes of interest directly into the preferred genetic background.

Much of crop domestication historically has involved increased yield and size (for example of fruit or seed), and this has resulted in parallel and often convergent selection upon traits and even genes of interest. Recent work by Dorian Fuller and colleagues used archaeological plant remains from around the world to examine the parallel acquisition of so-called “domestication syndrome traits” across both plant species and regions (Fuller et al. 2014). The authors found differences in the rate of evolution among domestication traits, but also saw remarkably similar rates across regions over periods spanning several centuries to millennia. This work was part of a special feature in PNAS on “the modern view of domestication,” in which 25 researchers from across five fields came together to both discuss progress being made in research on domestication and to identify key challenges for the future (Larson et al. 2014). The feature highlights the role of past domestication in shaping the variation in agricultural species we observe today and suggests future studies should address the role that the contemporary environmental and ecological context may have played in influencing selection on traits in the past.

Improved understanding of the evolutionary process as well as major technological advances means the pace of artificial selection has intensified and our ability to respond to changes in both the abiotic and biotic environment has improved greatly. Our ability to translate understanding of plant genetics and genomics into meaningful applications in crop science is discussed in a new piece by Pamela Ronald (Ronald 2014). The work emphasizes not only the great potential that marker assisted selection, genetic engineering, and genome editing hold in translational research, but also the great need to ensure such technologies benefit farmers in less well-developed countries. Of course the success of newly introduced agricultural varieties will depend on both the local environment and the subsequent evolution of other interacting species. As such, two new papers have focused on the importance of taking into account the evolutionary response of disease agents when guiding disease management practice in agriculture (Burdon et al. 2014; Zhan et al. 2014). For example, Jeremy Burdon and coauthors review the success of strategies such as stacking resistance genes, introducing partial or adult-only resistance, or using mixtures of host types to hinder pathogen evolution (Burdon et al. 2014). Similarly, Jiasui Zhan and collaborators discuss the importance of mimicking the spatial and temporal dynamics of natural host-pathogen coevolution when designing disease management strategies, and emphasize that resistance strategies with immediate short-term benefits are often the least durable in the long term (Zhan et al. 2014).

Given the rapid potential for adaptation, many predictions regarding pest or pathogen evolution can be directly tested in the laboratory in order to inform better disease management. Recent work by Julia Hillung and colleagues examined the adaptation of a plant RNA virus to various ecotypes of Arabidopsis thaliana in order to determine the specificity and consequences of evolution on one host to infectivity on another. They use experimental evolution to show rapid increases in infectivity and virulence on the host background in which the virus has been adapted, but also demonstrate that some host types select for viral populations that are more generally infective to other types (Hillung et al. 2014). These results are particularly intriguing in that they suggest manipulation of host types in an agricultural setting could predictably alter the outcome of pathogen evolution. Such rapid evolution is not restricted to the laboratory; evidence from the Western corn rootworm on maize crops indicates that the pest is evolving resistance to the toxins produced by genetically engineered plants that were introduced into production only in 2003 (Gassmann et al. 2014).

Importantly, the utility of evolutionary theory for agricultural practice is not limited to pest and pathogen interactions. The increasingly clear role of the microbiomes across the rhizosphere and phyllosphere suggest great potential for application of both community ecological and evolutionary thinking. Suzanne Donn and coauthors examined the changing soil microbiome of intensive wheat crops across years and found that, relative to soil in the absence of plants, rhizosphere communities changed substantially over time in the presence of plant roots and these temporal dynamics could be explained well based on the stage of plant development (Donn et al. 2014). Such knowledge about tightly coevolved plant-microbe interactions could help inform better management of soils and guide efforts to develop plant probiotics. Another attractive application of evolution to agriculture that has received recent attention is the incorporation of inclusive fitness theory. Toby Kiers and Ford Denison discuss ways in which artificial selection can be focused on improving cooperation among crop plants and the microbial symbionts with which they interact (Kiers and Denison 2014). For example, the authors suggest that the use of those crop types capable of imposing strong sanctions against “cheating” rhizobial bacteria strains (i.e. those that do not fix nitrogen as effectively) could increase the dominance of more mutualistic strains in the soil.

Overall, although artificial selection has been central to agricultural practice since its dawn, we are still constantly improving our ability to speed up the selective process, incorporate adaptation across heterogeneous environments, and allow for a more responsive management program in the face of coevolving enemies and mutualists. As such, there remains great promise in our ability to increase crop yield and decrease the use of pesticides and fertilizers through the application of evolutionary thinking.”

Literature cited

Burdon, JJ, LG Barrett, G Rebetzke, and PH Thrall 2014. Guiding deployment of resistance in cereals using evolutionary principles. Evolutionary Applications 7:609–624.

Donn, S, JA Kirkegaard, G Perera, AE Richardson, and M Watt. 2014. Evolution of bacterial communities in the wheat crop rhizosphere. Environmental Microbiology, doi: 10.1111/1462-2920.12452.

Fuller, DQ, T Denham, M Arroyo-Kalin, L Lucas, CJ Stevens, L Qin, RG Allaby et al. 2014. Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record. Proceedings of the National Academy of Sciences 111:6147–6152.

Gassmann, AJ, JL Petzold-Maxwell, EH Clifton, MW Dunbar, AM Hoffmann, DA Ingber, and RS Keweshan 2014. Field-evolved resistance by western corn rootworm to multiple Bacillus thuringiensis toxins in transgenic maize. Proceedings of the National Academy of Sciences 111:5141–5146.

Hillung, J, JM Cuevas, S Valverde, and SF Elena 2014. Experimental evolution of an emerging plant virus in host genotypes that differ in their susceptibility to infection. Evolution 68:2467–2480.

Kiers, ET, and RF Denison 2014. Inclusive fitness in agriculture. Philosophical Transactions of the Royal Society B: Biological Sciences369:20130367.

Larson, G, DR Piperno, RG Allaby, MD Purugganan, L Andersson, M Arroyo-Kalin, L Barton et al. 2014. Current perspectives and the future of domestication studies. Proceedings of the National Academy of Sciences 111:6139–6146.

Ronald, PC 2014. Lab to farm: applying research on plant genetics and genomics to crop improvement. PLoS Biology 12:e1001878.

Zhan, J, PH Thrall, and JJ Burdon 2014. Achieving sustainable plant disease management through evolutionary principles. Trends in Plant Sciences 19:570–575.

As I see it: the value of double blind peer review

When Michelle Tseng (founding editor of Evolutionary Applications) asked me many years back how I felt about double blind peer review, I was fairly agnostic. Wouldn’t most reviewers be able to guess anyway? Surely the system isn’t biased enough to warrant such an obstacle? How will reviewers know what sort of overlap the study has with other work the authors have published? And so forth. I am now changing my stance. And yes, this change is based on only an N of 1, so I would love to hear the thoughts and experiences of others who’ve gone the process lately (translation: comments welcome!! And no need to agree, of course).

A year and a half or so ago, I was contacted by Derek Lin, a previous undergraduate in John Thompson’s lab at UC Santa Cruz with whom I had collaborated on an experiment examining bacterial resistance against multiple phages. He was contemplating his next career move, and was considering both graduate school and a medical degree, but was currently enjoying his job as a teacher in the Bay Area. Derek also said he was interested in collaborating again and working on another paper, so we set up a time to Skype and compared ideas. After some brainstorming and looking around, Derek came up with the idea for a review article focusing on the human-associated pathogen, Helicobacter pylori, as he had been intrigued by a recent paper suggesting that the range of beneficial to pathogenic symptoms correlating with H. Pylori infection might be due to mismatched strains and hosts (see Kodamam et Al. 2014). I agreed that this would be an interesting topic to explore, and thus began the collaboration.

After over a year of research and back and forth of who knows how many drafts, Derek and I were ready to submit (but were both a bit nervous, as neither of us had ever worked on this particular pathogen before. Would the reviewers wonder why we thought we were in a position to write such a piece?). At the same time, I had an email from Craig Primmer, the Evolutionary Applications Reviews Editor, reminding me that the journal now had a special reviews and synthesis section. I thought to myself: everything’s coming up Milhouse! It was the perfect fit, as we had worked to take an evolutionary angle in reviewing the literature and to put forward some ways in which evolutionary theory could be applied to this topic. We submitted, and a little over a month later had our reviews back.

Okay… Here begins my conversion. Like many of you, I imagine, I am used to fairly patronising reviews that seem to always use the working assumption that I have not thought about alternative interpretations and hypotheses, do not have the expertise needed to write a paper, or am generally a numpty. These are always hard to read because, of course, I do have imposters syndrome and find putting my ideas and research out there into the public domain to be judged hard enough already. It takes me quite a while and a lot of work before I feel confident enough to submit a paper… So having negative reviews, especially when I find them unconstructive and occasionally just plain wrong, but worded strongly, is hard to swallow. I thought this was just how the review process worked, and that I needed tougher skin to stay in this field. This may still be the hard and fast truth, but I have just had the first seed of doubt planted.

When I got the recent reviews back from Evol Apps, I read through them and smiled. Not because they were overwhelmingly positive; they had some really useful criticisms and pointed out key gaps we needed to fill. Rather, my smile was due to the fact that they seemed to have been written with the underlying assumption that we knew what we were talking about (even referred to the piece as an “extensive and up-to-date review” – Rev 1, which “offers a welcome, balanced perspective” – Rev 2). How refreshing! I don’t know who the referees were, but I do know that I suggested 5 names in the field whom I’ve never met but seemed to be leading the way in H. Pylori research. After all, the point of this peer review process was to ensure we had not misrepresented or misunderstood the current state of the field. Later that day I was sharing this story at lunch where it was pointed out to me that the reviewers may have thought the authors were big shots in the field, or at least that they couldn’t rule this possibility out. Indeed, the real benefit of double blind peer review, to my mind, became obvious: the referees had to review the work on the science behind it, not the authors. I won’t speculate here as to whether the difference in the tone of these reviews from so many of my others comes down to a gender issue, or is simply due to my early(ish) career status, or is just a general phenomenon (although there are some empirical reasons to believe the first option may be true in some cases, e.g. here and here). All I can say is that this experience has made me much more likely to consider a journal with double blind peer review in the future. In part because I am a scientist, and don’t believe any study (or blog post) that is not based on properly replicated data.

More resources of interest:

Kodaman, Nuri, et al. “Human and Helicobacter pylori coevolution shapes the risk of gastric disease.” Proceedings of the National Academy of Sciences 111.4 (2014): 1455-1460.

Disease evolution and ecology across space

For this month’s Evolutionary Application Research Highlight, I explored:

Disease evolution and ecology across space

“How infectious disease spreads both from individual to individual and across a landscape will depend upon many inter-related factors, including the genetic composition of host and pathogen populations, the pathogen transmission rate, host density, population connectivity, and the evolutionary response of both host and pathogen over time. As such, the study of infectious disease straddles a number of fields and approaches. A key advance in the field has come from incorporation of spatial structure into both theoretical and empirical studies of the epidemiology, evolution and ecology of disease. Researchers taking this approach were recently brought together for a workshop entitled “Spatial evolutionary epidemiology” in Montpellier, France organized by Sébastien Lion and Sylvain Gandon from the Centre d’Écologie Fonctionnelle et Évolutive (CEFE). The workshop spanned topics from infection genetics in Daphnia, to cooperation and conflict in microbial populations, to the utility of spatial epidemiological models for designing better crop planting strategies, and emphasized the broader need to think about the importance of spatial heterogeneity in order to predict the spread and evolution of pathogens.

The recent literature includes a number of studies at the cutting edge of this field. For example, a model based on interactions between bacteria and their bacteriophage parasites by Ben Ashby and collaborators has demonstrated the importance of spatial structure in shaping the breadth of host resistance. They find that hosts and parasites from spatially structured populations should be less constrained by costs associated with ‘generalism’ than those from well-mixed populations, and therefore that spatial structure is likely to increases the breadth of host resistance/parasite infectivity, especially when this increased breadth carries a significant fitness cost (Ashby et al. 2014). Another potential effect of spatial structure on host-parasite interactions is through changing rates of infection, as opposed to infection range. A recent empirical study of bacteria and phages by Pavitra Roychoudhury and coauthors experimentally evolved phages on spatially structured agar plates for 550 phage generations and found increased phage fitness was associated with a mutation conferring slower phage adsorption rate to bacterial host cells. This is in line with theoretical work suggesting spatial structure may lead to reduced parasite infectivity (e.g. Boots and Sasaki 1999) and builds upon previous empirical work from non-microbial systems in support of this theory. Furthermore, the authors develop a system-specific, spatially explicit model to explain how low attachment probability might lead to increased phage fitness through higher plaque density as a result of trade-offs between phage diffusion and adsorption (Roychoudhury et al. 2014).

Studying spatial structure in natural populations is of course a more challenging goal, as one loses the ability to control for the type of heterogeneity across which a study is performed. Despite these potential limitations, recent work by Jussi Jousimo, Ayco Tack, and collaborators examined the impact of connectivity among over 4000 populations of the plant, Plantago lanceolata, on resistance to a co-occurring fungal pathogen, Podosphaera plantaginis, over a 12-year period. Using this large spatial dataset, the authors were able to explain the unexpected observation that highly connected populations typically showed lower pathogen prevalence and higher pathogen extinction than more isolated populations. They demonstrate that hosts from highly connected populations are in fact more resistant to the pathogen than those from more isolated populations, most likely as a result of stronger parasite-mediated selection over time (Jousimo et al. 2014). This work highlights the importance of incorporating host and pathogen evolution into models predicting the spread of disease across space, and emphasizes the potential for differential parasite-mediated selection across a host metapopulation.

Finally, work by David Rasmussen and colleagues has demonstrated the strength of incorporating spatial structure into phylodynamic (coalescent) approaches to inferring past disease dynamics from genealogies (Rasmussen et al. 2014). Motivated to explain the common discrepancy between phylodynamic inferences and those made from hospital records, the authors apply their new method to a case study of mosquito-borne dengue virus in southern Vietnam. They demonstrate that a spatially structured susceptible-infected-recovered (SIR) model resulted in similar patterns to those seen from the hospitalization data, including large seasonal fluctuations in disease, and that the incorporation of ecological complexity into coalescent models increases the accuracy of inferences to disease demography.

Overall, these recent studies demonstrate the added value of incorporating spatial structure into epidemiological, phylodynamic, and evolutionary/ecological models of infectious disease. Although the current body of theory on the topic offers clear predictions for how spatial structure might influence disease evolution and spread, there remains a paucity of empirical and observational studies testing these key ideas. Moving forward, the more we understand about these eco-evolutionary feedbacks, the better we will be able to manage emerging disease in natural, agricultural and human populations.”

References cited:

Ashby, B., Gupta, S. and A. Buckling. 2014. Spatial Structure Mitigates Fitness Costs in Host-Parasite Coevolution. The American Naturalist 183:E64-E74.

Boots, M. and A. Sasaki. 1999. ‘Small worlds’ and the evolution of virulence: infection occurs locally and at a distance. Proceedings of the Royal Society of London. Series B: Biological Sciences 266:1933-1938.

Jousimo, J., Tack, A. J., Ovaskainen, O., Mononen, T., Susi, H., Tollenaere, C. and A. L. Laine. 2014. Ecological and evolutionary effects of fragmentation on infectious disease dynamics. Science 344:1289-1293.

Rasmussen D. A., Boni M. F. and K. Koelle. Reconciling phylodynamics with epidemiology: the case of dengue virus in southern Vietnam. Molecular biology and evolution 31:258-271.

Roychoudhury, P., Shrestha, N., Wiss, V. R. and S. M. Krone. 2014. Fitness benefits of low infectivity in a spatially structured population of bacteriophages. Proceedings of the Royal Society B: Biological Sciences 281:20132563.

The utility of model system research for applied evolutionary questions

For this month’s Evolutionary Application’s research highlights, I look at recent work exemplifying the use of model systems in addressing questions of applied interest:

“Our ability to apply evolutionary theory is necessarily limited by our understanding of natural systems. Unfortunately, given the finite amount of researcher time and funding, we are faced with a trade-off between studying many systems shallowly and studying fewer systems in more depth. The handful of ‘model systems’ that have emerged thus far act as workhorses across disciplines, allowing for a more complete understanding of each system in fields from molecular genetics and evolution, to development, to physiology, to whole organism biology and ecology. However, given how few such model systems exist, it remains to determine how generalizable they are to less well-studied species and the natural world as a whole.

Among the model systems offering a wealth of insight across fields are the yeast Saccharomyces cerevisiae, the nematode Caenorhabditis elegans, the fly Drosophila melanogaster, and the plant Arabidopsis thaliana. These systems, among others, became models due in part to their ease of use in the lab, relative ubiquity, short generation time, and, in part, due to chance. The breadth of knowledge gained from the study of these and other model systems has driven progress across fields and has allowed for multidisciplinary research and cross talk among researchers that may not have otherwise come together. More recently, there has been a renewed interest in examining these systems back in the field and expanding the work to closely related species to determine whether our knowledge from the laboratory is broadly applicable in nature.

In terms of real life application, the model yeast Saccharomyces cerevisiae has the clearest relevance given its use in fermentation processes, the evolutionary history of which has been recently reviewed by Dashko et al. (2014). However, the tremendous toolbox of genetic techniques that is available has also made S. cerevisiae a central player in answering more basic questions in biology. Recent work by Serero et al. (2014) used a range of molecular approaches to examine mutation accumulation of both wild type and mutator strains of S. cerevisiae (i.e., strains with deficiencies in so-called ‘genome maintenance’ genes). By comparing the genome-wide mutational landscape across mutator types, they demonstrate strain-specific and complex effects of mutagenesis on chromosomal structure, mutations, and aneuploidy. The study emphasizes the tremendous diversity of the mutational landscape and provides an approach for examining genomic variation during clonal evolution, as occurs for example during tumor development in cancer.

The nematode Caenorhabditis elegans has acted as powerful model system for studying evolution due in part to the ability of researchers to freeze and resurrect individuals. A recent review by Gray and Cutter (2014) highlights the great potential the system still holds in exploring mutational processes, mating system and life-history evolution, and host-pathogen coevolution. For example, a new study by Sikkink et al. (2014) has examined the importance of phenotypic plasticity in adapting to extreme environments using the outcrossing sister species C. remanei. They experimentally selected for worms that were able to better withstand heat shock during development and found both increased tolerance of heat shock and altered phenotypic plasticity when reared in a novel environment, emphasizing both the role evolution can play in shaping plasticity and the importance of plasticity in allowing adaptation to novel and/or extreme environments.

Research on the fruitfly, Drosophila melanogaster, has paved the way for our understanding of genetics but has also been central to addressing questions regarding the impact of symbionts and pathogens on eukaryotic fitness (reviewed in Fauvarque 2014). Recent work by Versace et al. (2014) experimentally evolved populations of D. melanogaster infected with Wolbachia from multiple clades under hot or cold environments for 37 generations to test for changes in symbiont composition across environments. They discovered rapid increase in infection rates across replicate populations and treatments, suggesting a strong fitness advantage to hosts carrying the symbiont, as well as striking shifts in the composition of the Wolbachia community under cold, but not hot, environmental conditions. Studies of Drosophila have also been useful in understanding pest emergence and evolution. For example, a study by Atallah et al. (2014) has compared those Drosophila species that feed primarily on rotting fruit with those species that have switched to live fruit. Through morphological comparisons of the pest D. suzukii with its close relatives, the authors propose an evolutionary model to explain the modification of the ovipositor that allows puncture of susceptible fruits, demonstrating the utility of an evolutionary framework for addressing questions of pest emergence and management.

Finally, research focused on Arabidopsis thaliana plants has offered key insights to the genetics of plant adaptation, plant development, and plant–pathogen interactions. Two recent studies have examined natural populations of Arabidopsis to better understand the plant’s ability to move beyond current range limits. First, Wolfe and Tonsor (2014) took advantage of a natural elevational gradient of temperature and precipitation to examine plant adaptation to increase heat and drought. By exposing 48 lineages from across the natural gradient to increasing temperature and decreasing precipitation, they show that 10 of the 12 traits measured differ across the elevations and that populations from the low elevations are most fit in the face of increased heat and drought. In particular, these plants have faster bolting and earlier fruit ripening than those from high elevations, suggesting adaptation toward avoidance of spring heat and drought. In another study, Griffin and Willi (2014) examined natural populations of another Arabidopsis species, A. lyrata, across North America to test whether self-fertilization is more common at the edge of the species range, where effective population size is likely to be lower. Using population surveys and microsatellite markers, they were able to demonstrate at least seven independent transitions from outcrossing to selfing, all of which occurred at the edge of the species range where diversity was lower. Understanding such evolutionary shifts toward self-fertilization offers important insight to the ability of a species to expand its range and potentially to become invasive.

These recent studies highlight the great potential model systems hold for applying evolutionary theory in large part due to their amenability to laboratory conditions and genetic/genomic tools. Overall, it is clear that the knowledge gained from the study of model systems has been greater than the sum of its parts, but the generalizability of such knowledge to nonmodel systems, especially when it comes to translational research, remains an open question.”

References cited:

Atallah, J., L. Teixeira, R. Salazar, G. Zaragoza, and A. Kopp 2014. The making of a pest: the evolution of a fruit-penetrating ovipositor in Drosophila suzukii and related species. Proceedings of the Royal Society B: Biological Sciences 281:20132840.

Dashko, S., N. Zhou, C. Compagno, and J. Piškur 2014. Why, when and how did yeast evolve alcoholic fermentation? FEMS Yeast Research, doi: 10.1111/1567-1364.12161.

Fauvarque, M. O. 2014. Small flies to tackle big questions: assaying complex bacterial virulence mechanisms using Drosophila melanogaster. Cellular Microbiology 16:824–833.

Gray, J. C., and A. D. Cutter 2014. Mainstreaming Caenorhabditis elegans in experimental evolution. Proceedings of the Royal Society B: Biological Sciences 281:20133055.

Griffin, P. C., and Y. Willi 2014. Evolutionary shifts to self-fertilisation restricted to geographic range margins in North American Arabidopsis lyrata. Ecology Letters 17:484–490.

Serero, A., C. Jubin, S. Loeillet, P. Legoix-Né, and A. G. Nicolas 2014. Mutational landscape of yeast mutator strains. Proceedings of the National Academy of Sciences 111:1897–1902.

Sikkink, K. L., R. M. Reynolds, C. M. Ituarte, W. A. Cresko, and P. C. Phillips 2014. Rapid evolution of phenotypic plasticity and shifting thresholds of genetic assimilation in the nematode Caenorhabditis remanei. G3: Genes – Genomes – Genetics 4:1103–1112.

Versace, E., V. Nolte, R. V. Pandey, R. Tobler, and C. Schlötterer 2014. Experimental evolution reveals habitat-specific fitness dynamics among Wolbachia clades in Drosophila melanogaster. Molecular Ecology 23:802–814.

Wolfe, M. D., and S. J. Tonsor 2014. Adaptation to spring heat and drought in northeastern Spanish Arabidopsis thaliana. New Phytologist 201:323–334.


The role of the microbiome in shaping evolution

For this month’s Evolutionary Applications research highlights I look at the role of the microbiome in shaping evolution:

“Over the past century, the study of genetics has revolutionized our understanding of life on earth. Our knowledge of trait heritability from parent to offspring has been central to predict the trajectory of evolution, studying disease, and successful breeding of crops and animals. The field of genetics continues to grow in leaps and bounds due to next-generation sequencing, metagenomic approaches, genetic engineering, a better understanding of epigenetics, and, most recently, the creation of synthetic chromosomes (Annaluru et al. 2014). Despite these advances, however, there is still an active debate regarding how much variation in phenotype is explained by nature (the genome) versus nurture (the environment; recently reviewed in Lynch and Kemp 2014). In addition, it is increasingly apparent that a significant proportion of the so-called missing heritability may be explained by host-associated microbial communities, the microbiome.

The microbiome of eukaryotes has been associated with traits ranging from disease susceptibility to digestion to behavior and even holds the potential to drive speciation (Brucker and Bordenstein 2013). This rapidly growing field already has its own journal (‘Microbiome,’ established in 2013) and has been the focus of a recent special issue of Microbial Ecology on ‘Nature’s microbiome’ (Russell et al. 2014), in which 28 research groups present new ideas and data on the composition and function of the microbiome and on how microbe–microbe and host–microbe interactions might shape evolution. Among the recent headlines, we have seen a role for soil-associated microbes in creating the taste of particular wine vintages (Bokulich et al. 2014) and good evidence that immune defense is modulated both directly and indirectly by our microbiota (recently reviewed in Abt and Pamer 2014). In addition, work by Maggie Wagner and colleagues on a wild relative of Arabidopsis has uncovered the key role of soil microbiota both in shaping flowering time and in influencing the intensity of selection on flowering time (Wagner et al. 2014).

Given the complexity of studying the human microbiome, much of our current understanding comes from work carried out on germ-free mice. A recent study by Jeremiah Faith and coauthors, which introduced 94 bacterial consortia of diverse sizes chosen at random from human fecal samples, was able to uncover key roles of the microbiota in inflammatory responses, obesity and variation in metabolites in mouse hosts (Faith et al. 2014). Determining how human microbiomes are shaped and how they may have coevolved with the population requires very large datasets to account for the great variation in diet, geography, race, and lifestyle among individuals. However, recent insight into how microbiota may have changed due to urbanization has come from Stephanie Schnorr and colleagues, who sequenced the microbiome of individuals from the Hadza hunter–gatherer community in Tanzania (Schnorr et al. 2014). They find evidence suggesting that, relative to individuals from urban communities in Italy, the Hadza microbiota is typically more species rich and lacks the typically common Bifidobacterium. Of course, as recently discussed by Eva Boon and collaborators, the importance of variation in microbiota composition among individuals and populations is less about what species are there than it is about what genes are there (Boon et al. 2014). This is both because of redundancy in function among microbial species and also due to the ability of bacteria to horizontally transmit genetic material among genomes, such that one population can readily evolve a new function simply by acquiring the necessary genes.

Given the current open questions regarding the function and composition of human microbiota, we are not yet at the stage of developing artificial communities as treatment for disease. However, in extreme cases of Clostridium difficile infection, doctors have been turning to fecal transplants as a way of resetting the microbiome of their patients with remarkably high success rates. Susana Fuentes and colleagues recently tracked changes in the microbiome of patients before, during and after such transplants and found a marked and long-lasting increase in microbial diversity after the transplant (Fuentes et al. 2014). It remains to be determined whether success rate is affected by interactions between the host genotype and the transplant microbiota, but we can look to data from other organisms for such clues. For example, Marie-Lara Bouffaud and coauthors report a significant relationship between rhizobacterial communities and genetic distance of their plant hosts, and this relationship held when looking only at single bacterial species (Bouffaud et al. 2014). On the other hand, work by Julie Reveillaud and colleagues found no clear signature of host relatedness in explaining the microbiota associated with coral hosts, although their data do suggest species-specific microbiota communities even across geographically distant deep sea populations (Reveillaud et al. 2014).

Another potential application of microbiome research is the use of ‘prebiotics,’ particular dietary fibers, to manipulate the composition of the microbiota. Amandine Everard and coauthors tested the impact of prebiotic treatment on mice that were fed high-fat diets and found that the differing composition in microbiota of treated mice acted to counteract inflammation and metabolic disorders induced by the high-fat diet (Everard et al. 2014). However, to fully translate the burgeoning microbiome data into practical applications, such as the use of pre- or probiotics to prevent/treat disease or to alter organismal phenotype in a predictable way, we need to untangle the complexity of social interactions among microbes more generally. This idea has been highlighted by Helen Leggett and coauthors, who review the wide range of ways in which microbes interact within their eukaryotic hosts (Leggett et al. 2014). The review emphasizes that better insight into microbe–microbe social evolution, both within and between species, will be central to better predicting the evolution of virulence, drug resistance, and the spread of infectious disease. The idea of harnessing information about social interactions, including those between microbes, to design novel treatments has been coined ‘Hamiltonian medicine’ and recently conceptualized by Bernard Crespi and colleagues (Crespi et al. 2014).

Together, the wealth of new data emphasizes that microbiota play central roles in shaping the health, ecology, and evolution of their hosts. The application of microbiota research is currently hindered by the complexity of the interactions (both among microbes and between the microbiota and the host), but the potential for application of this knowledge seems limitless.”

References cited:

Abt, M. C., and E. G. Pamer. 2014. Commensal bacteria mediated defenses against pathogens. Current Opinion in Immunology 29:16–22.

Annaluru, N., H. Muller, L. A. Mitchell, S. Ramalingam, G. Stracquadanio, S. M. Richardson, and M. E. Linder. 2014. Total synthesis of a functional designer eukaryotic chromosome. Science 344:55–58.

Bokulich, N. A., J. H. Thorngate, P. M. Richardson, and D. A. Mills 2014. Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate. Proceedings of the National Academy of Sciences 111:E139–E148.

Boon, E., C. J. Meehan, C. Whidden, D. H. J. Wong, M. G. Langille, and R. G. Beiko 2014. Interactions in the microbiome: communities of organisms and communities of genes. FEMS Microbiology Reviews 38:90–118.

Bouffaud, M. L., M. A. Poirier, D. Muller, and Y. Moënne-Loccoz 2014. Root microbiome relates to plant host evolution in maize and other Poaceae. Environmental Microbiology, DOI: 10.1111/1462-2920.12442.

Brucker, R. M., and S. R. Bordenstein 2013. The hologenomic basis of speciation: gut bacteria cause hybrid lethality in the genus Nasonia. Science 341:667–669.

Crespi, B., K. Foster, and F. Úbeda 2014. First principles of Hamiltonian medicine. Philosophical Transactions of the Royal Society B: Biological Sciences 369:20130366.

Everard, A., V. Lazarevic, N. Gaïa, M. Johansson, M. Ståhlman, F. Backhed, and P. D. Cani 2014. Microbiome of prebiotic-treated mice reveals novel targets involved in host response during obesity. The ISME Journal, DOI:10.1038/ismej.2014.45.

Faith, J. J., P. P. Ahern, V. K. Ridaura, J. Cheng, and J. I. Gordon 2014. Identifying gut microbe–host phenotype relationships using combinatorial communities in gnotobiotic mice. Science Translational Medicine 6:220ra11.

Fuentes, S., E. van Nood, S. Tims, I. Heikamp-de Jong, C. J. ter Braak, J. J. Keller, and W. M. de Vos 2014. Reset of a critically disturbed microbial ecosystem: faecal transplant in recurrent Clostridium difficile infection. The ISME Journal, DOI:10.1038/ismej.2014.13.

Leggett, H. C., S. P. Brown, and S. E. Reece 2014. War and peace: social interactions in infections. Philosophical Transactions of the Royal Society B: Biological Sciences 369:20130365.

Lynch, K. E., and D. J. Kemp 2014. Nature-via-nurture and unravelling causality in evolutionary genetics. Trends in Ecology & Evolution 29:2–4.

Reveillaud, J., L. Maignien, A. M. Eren, J. A. Huber, A. Apprill, M. L. Sogin, and A. Vanreusel 2014. Host-specificity among abundant and rare taxa in the sponge microbiome. The ISME Journal, DOI:10.1038/ismej.2013.227.

Russell, J. A., N. Dubilier, and J. A. Rudgers 2014. Nature’s microbiome: introduction. Molecular Ecology 23:1225–1237.

Schnorr, S. L., M. Candela, S. Rampelli, M. Centanni, C. Consolandi, G. Basaglia, and A. N. Crittenden 2014. Gut microbiome of the Hadza hunter-gatherers. Nature Communications, DOI:10.1038/ncomms4654.

Wagner, M. R., D. S. Lundberg, D. Coleman-Derr, S. G. Tringe, J. L. Dangl, and T. Mitchell-Olds 2014. Natural soil microbes alter flowering phenology and the intensity of selection on flowering time in a wild Arabidopsis relative. Ecology Letters 17:717–726.

Guest post by Sean Meaden

Do we need to watch what we spray? A summary of our recent review on the potential dangers of phage biopesticides.

Guest post by Sean Meaden, PhD student at University of Exeter working on phage-Pseudomonas syringae-plant host interactions.

It seems barely a week goes by without mention of the dangers of antibiotic resistance in popular news stories. Just last month the WHO called antibiotic resistance a ‘global threat’ in a well-publicized press release [1]. Whilst this might be slightly good news for those of us soon to be looking for post-doc positions in the field of microbiology, it certainly isn’t good for public health. The need for alternatives is pressing and it seems there just isn’t the financial incentive for large pharma companies to develop new drugs. Better stewardship of the compounds we already use is crucial, but so is exploring alternative options and exploiting our knowledge of microbial ecology and evolution.

One alternative strategy, known as ‘phage therapy,’ is the utilization of viruses that infect bacteria to decrease density of specific bacterial populations. In nature, phages have been estimated to kill as much as 20% of global bacterial populations each day [2]. As many reviews have noted, the use of phages is nothing new and has been a successful ongoing industry in Georgia and the former Soviet Union since the 1920s [3]. The aim of phage therapy is to find or create a phage that specifically infects a pathogenic strain of bacteria, culture it in the lab and turn it into a useable end product that can be ingested, applied topically or sprayed onto crops (as a biopesticide).

Phage cocktail production
Typical production of phage biopesticide. Reprinted from Meaden S and Koskella B (2013) Exploring the risks of phage application in the environment. Front. Microbiol. 4:358. doi: 10.3389/fmicb.2013.00358

The potential benefits of phage therapy are huge. However, given the negative consequences of improperly managed antibiotic use, we should be careful not to make the same mistakes again. It strikes me that a common theme in the phage therapy literature is ‘It’s OK, phages are naturally occurring,’ implying (in my mind at least) ‘What can go wrong?’ The same argument could have been made about antibiotics. After all, they too are naturally occurring compounds produced for microbial battles among themselves, and are likely to have been around for billions of years [4]. In our recent review [5] we explored the possible negative consequences of phage therapy to assess whether we are likely to make the same mistakes as we did with our often poor stewardship of antibiotics. Below is my summary of the main arguments we outline in the paper. On the whole I think the need for alternatives to antibiotics is huge, and by being prudent with our use of phages as a replacement (or synergistic treatment), and critically evaluating negative consequences, we will be better placed to use phage therapy successfully.

1) Implications of evolved resistance. This fairly obvious point is probably most comparable to antibiotic usage. We know that phages and bacteria undergo arms races of resistance and counter-resistance [6]. If the pathogens that are causing disease evolve resistance to the phages that we use as therapeutic agents, our cure becomes defunct and we have to go back to the drawing board. Finding new infective phages shouldn’t be too hard, but the process of turning them into a usable product that has passed regulatory hurdles is likely to be lengthy. This is a parallel problem to antibiotic production- there must be novel antibiotic compounds in the soil under our feet, but turning them into a lifesaving drug is the tricky part. A solution proposed by a group working on burn patients in Belgium is to create a reactive, cottage-industry style phage therapy centre that quickly screens for phage infectivity from a ready-made library, rather than a single product for widespread consumption (and most desirable to big pharma) [7]. Whilst this approach is great for pathogens that are readily culturable in the lab it might be more difficult for less tractable organisms.

2) Phage mediated attenuation of bacterial resistance. This argument seems to pop out of the literature as ‘OK, the bacteria might evolve resistance, but that won’t matter because resistance is costly so their virulence will be attenuated’. In a few cases this certainly does seem to be the case, for example Filippov et al. found reduced virulence of Yersinia pestis in mice (in other words mouse plague was less deadly when phage resistance had evolved; 8). In other cases, phage resistance actually had the opposite effect, making Pseudomonas aeruginosa more virulent in vitro, and rightly highlighting the need for caution in selecting phages for treatment [9]. Thus, for this argument to be used informatively we need much more data from a variety of systems and under more natural conditions or full-scale trials.

3) Agricultural cross-over. A consistent criticism of global policy on antibiotic stewardship is the use of antibiotics in agriculture. The addition of antibiotics at sub-therapeutic levels is great for feed efficiency and increasing yields. Given the increasing demand for protein in the global diet this issue isn’t trivial. However, it is likely that such practices increase levels of antibiotic resistance, and the bi-directional exchange of resistance genes from farm to community should be worrying. If phage therapy becomes more commonplace in agricultural settings could we see the same effect? Managed carefully, cross-resistance between agricultural and clinical phage therapeutics shouldn’t be a problem, especially given the typically (but not exclusively) high host specificity of phages. But I do think it’s worth acknowledging the potential for interference in order to prevent repeating mistakes of the past.

4) Horizontal gene flow from phage application. Phages are so good at transferring genes among bacterial cells that we use them in the lab to do just that. It’s certain that this horizontal exchange goes on in the environment, so we must ask: if we pump out unnaturally high volumes of phages (especially those with broader host ranges) into the environment, how likely are these phages to move genes around. This is especially problematic when the genes being swapped encode antibiotic resistance, toxins, or virulence factors. In this case, our attempt at a cure could actually make things much worse. We know that phage-mediated gene-transfer has played a part in cholera epidemics [10] and we should be careful about facilitating the spread of other unwanted bacterial traits.

5) Impacts on natural communities. The importance of a ‘healthy’ microbiome is constantly espoused (so much so that Jonathan Eisen has produced an award for ‘Overselling the Microbiome;’ 11). Although the direct effects of a healthy microbiota are still predominantly correlational, minimizing the disruption to a community whilst removing a pathogenic species must surely be the ultimate goal. Phages could hold great potential in this regard as they tend to be fairly specific in their host range (so could act more like snipers) relative to antibiotics (which act more like indiscriminate hand-grenades). On the other hand, artificially high volumes of phages could have unexpected effects on microbial processes, particularly in an agricultural environment. To my knowledge the effects of adding high titres of phages to microbial communities in the environment remains unexplored.

6)   Unpredictability of infection kinetics. This issue is an exciting one- the pioneer of phage therapy, Felix d’Herelle, is spuriously quoted as stating that immunity is contagious as well as the cure [12]. He might have been wrong about the biology but the premise that the cure is transmissible certainly seems possible with replicating phages. The downside is that it makes it hard to predict the persistence of phages in the environment. Unlike an antibiotic with a known half-life, a phage therapy product could continue to persist and replicate in the environment indefinitely. Even at low densities this raises an ethical question of something being uncontainable.

Recently, the ethical imperative of using phage therapy has been stated [13] and there is clearly a need for alternative strategies to antibiotic drugs. All of the concerns raised in our review are addressable and shouldn’t preclude the use of phage therapy in a clinical or agricultural setting. Moreover, all of the questions we have raised are readily answerable given the advances of microbial genomics over the last decade. We just need more data!!

If this summary piqued your interest, check out the whole article here, and open access:


  2. Suttle, C. A. 1994 The significance of viruses to mortality in aquatic microbial communities. Microb. Ecol. 28, 237–43.
  3. Kutateladze, M. and Adamia, R. 2008 Phage therapy experience at the Eliava Institute. Med Mal Infect 38, 426–430.
  4. D’Costa, V. M. et al. 2011 Antibiotic resistance is ancient. Nature 477: 457–461.
  5. Meaden, S., & Koskella, B. (2013). Exploring the risks of phage application in the environment Frontiers in Microbiology, 4 DOI: 10.3389/fmicb.2013.00358
  6. Buckling, A. and Rainey, P.B. 2002. Antagonistic coevolution between a bacterium and a bacteriophage. Proc. R. Soc. Lond. Ser. B, 269: 931–936
  7. Pirnay, J.-P. et al. 2011. The phage therapy paradigm: pret-a-porter or sur-mesure? Pharmaceutical Research 28:934–937.
  8. Filippov, A. A. et al. 2011 Bacteriophage-resistant mutants in Yersinia pestis: identification of phage receptors and attenuation for mice. PLoS ONE 6, e25486
  9. Hosseinidoust, Z. et al 2013 Evolution of Pseudomonas aeruginosa virulence as a result of phage predation. Appl Environ Microbiol 79: 6110–6116
  10. Waldor, M. K. and Mekalanos, J. J. 1996 Lysogenic conversion by a filamentous phage encoding cholera toxin. Science 272:1910–1914.
  12. d’Herelle, F. 1924 Immunity in Natural Infectious Disease, Williams & Wilkins
  13. Verbeken, G. et al. 2014 Taking bacteriophage therapy seriously: a moral argument. BioMed Research Int. 2014: Article ID 621316, 8 pages.