Combining experimental evolution with next-generation sequencing: a powerful tool to study adaptation from standing genetic variation

Evolve and resequence (E&R) is a new approach to investigate the genomic responses to selection during experimental evolution. By using whole genome sequencing of pools of individuals (Pool-Seq), this method can identify selected variants in controlled and replicable experimental settings. Reviewing the current state of the field, we show that E&R can be powerful enough to identify causative genes and possibly even single-nucleotide polymorphisms. We also discuss how the experimental design and the complexity of the trait could result in a large number of false positive candidates. We suggest experimental and analytical strategies to maximize the power of E&R to uncover the genotype–phenotype link and serve as an important research tool for a broad range of evolutionary questions.Experimental evolution has a long tradition in biology (Garland and Rose, 2009). By exposing an evolving population to conditions chosen by the researcher, it is possible to study the response to this selection regime. A recent review highlighted the broad range of applications that have been investigated with this methodology and concluded that the breadth of research questions is only limited by the creativity of the experimenter (Kawecki et al., 2012). In addition to the great diversity of experimental designs, experimental evolution provides a unique advantage compared with other evolutionary analyses: the ability to replicate an experiment under identical conditions. Through this replication, experimenters are able to distinguish between stochastic and deterministic effects. Until recently, experimental evolution has mainly focused on phenotypes, sometimes combined with the analysis of a small number of markers (see, for example, Nuzhdin et al., 1993; Teotonio et al., 2009). In the wake of the latest sequencing technologies and the ongoing drop in DNA sequencing costs, however, the ultimate goal to connect the phenotypic response to the underlying genetic changes during an experimental evolution study has now come within reach.Depending on the starting population, two conceptually different approaches of experimental evolution can be distinguished. Either the experiment starts from a genetically homogeneous (invariable) population or from a polymorphic population. In the first approach, adaptation occurs through the accumulation of new beneficial mutations during the experiment (Elena and Lenski, 2003). These experiments therefore require very large population sizes and many generations to ensure a sufficient mutation supply and are thus largely restricted to microorganisms. Alternatively, experiments starting with a polymorphic population do not require novel mutations as selection can act on beneficial alleles that are already present at the beginning of the experiment. Given the massive genetic variation that is present in the starting population, the key challenge for this approach is distinguishing between selected and neutral variants. Neither randomly selected markers nor whole genome sequencing of a few representative individuals can provide sufficient information about the true target(s) of selection. Rather, genome-wide polymorphism data are needed.As whole genome sequencing is still not feasible for large numbers of individuals, experimental evolution studies starting from polymorphic base populations rely on a modified next-generation sequencing approach. Rather than sequencing individuals separately, DNA of multiple individuals from a population are sequenced together (Pool-Seq). This method is more cost effective than sequencing of individuals (Futschik and Schlötterer, 2010) and yields highly accurate genome-wide allele frequency estimates (reviewed in Rellstab et al., 2013; Schlötterer et al., 2014). The combination of experimental evolution with Pool-Seq is also known as Evolve and Resequence (E&R; Turner et al., 2011; Figure 1). Here, we review the state of the art of whole genome polymorphism analysis in experimental evolution studies relying primarily on segregating variation in the starting population.Open in a separate windowFigure 1Overview of E&R studies. (a) A population of flies is exposed for 60 generations to ultraviolet (UV) radiation (purple arrows). We assume here, for the sake of illustration, that darker pigmentation is beneficial in high UV environments, whereby darker flies will increase in frequency. (b) At the genotypic level, the allele frequency of the causative allele (dark brown) will increase, more so than hitchhiking variants (dark gray background) that will be recombined onto other backgrounds (breaks between dark and light gray background). (c) The allele frequencies of the starting population and the selected population are measured with Pool-Seq. (d) Causative variants can be identified by contrasting the allele frequencies between base and selected population and visualized with Manhattan plots. A full color version of this figure is available at the Heredity journal online.In many experimental evolution studies, researchers select for a well-defined trait in a controlled environment. This assures that both the phenotypic and the underlying genomic response are triggered either directly or indirectly by the selection regime applied during the experiment. Thus, E&R studies provide a complementary approach to genome-wide association studies (GWASs) and linkage mapping experiments as strategies to connect genotype and phenotype.     

Spatial Variation in δ15N and δ13C Isotopes in the San Juan River, New Mexico and Utah: Implications for the Conservation of Native Fishes

Synopsis Spatial patterns of resource use by small-bodied fishes in the San Juan River were examined using stable isotopes. Using δ¹⁵N of fishes as an index of trophic position, our data suggest both native and non-native fishes primarily consumed macro-invertebrates. The δ¹³C of these fishes further suggested a detritus-based food web, from which most species fed on chironomids in low-velocity habitats. A two-way ANOVA revealed a significant interaction between trophic level of fish species and longitudinal position in the river. This interaction was primarily attributed to a decline in trophic level of non-native red shiner Cyprinella lutrensis, relative to other species, in upstream reaches of the river. In addition, ANCOVA results suggest trophic position of fishes was dependent on channel type (primary vs. secondary), as there was less variability in resource use in secondary channels. These data provided a spatial framework of trophic interactions that can be used to predict the outcome of management actions. Overall, we confirmed high overlap in resource used between native and non-native fishes. However, spatial variation in trophic interactions both longitudinally and laterally in the river present a challenge to resource managers attempting to managing entire river systems.     

Lipase-mediated Resolution of Branched Chain Fatty Acids

Branched chain fatty acids (BCFAs) are fatty acids substituted with alkyl groups. Many of them are chiral and therefore occur in two enantiomeric forms. This review describes their occurrence in Nature, their biosynthesis, their properties as flavours, and their enzymatic kinetic resolution. Many lipases are able to separate the enantiomers of BCFAs, in hydrolysis, esterification or transesterification reactions. Very often, the stereoselectivity of these reactions is remarkably high, even when the chiral carbon atom is remote from the carboxylic acid group.     

Effects of citronella oil (Cymbopogon winterianus Jowitt ex Bor) on Spodoptera frugiperda (J. E. Smith) midgut and fat body

The armyworm, Spodoptera frugiperda, is the principal pest of corn in Brazil. Control is achieved primarily by synthetic insecticides, which cause problems for the agro-ecosystem. Alternative methods of control are under investigation and citronella (Cymbopogon winterianus) essential oil appears to be a promising agent. We investigated the effects of citronella oil using histological, histochemical and immunohistochemical methods. The midgut of larvae treated with citronella exhibited altered epithelium including cytoplasmic protrusions, columnar cell extrusion, pyknotic nuclei, and increased periodic acid-Schiff positive granules. Regenerative cells in the epithelium of the midgut increased in number, which facilitated subsequent regeneration of this tissue. After exposure to citronella, trophocytes, the principal cell type of the fat body, possessed enlarged vacuoles and mitotic bodies, and contained reduced amounts of glycogen, lipid, and protein. Citronella oil caused morphological changes of the midgut and reduction of stored resources in the fat body, which may adversely affect insect reproduction and survival.     

Reactor Design for the Novozym 435 ® -catalysed Enantioselective Esterification of 4-methyloctanoic Acid

The bench scale Novozym 435 ® catalysed esterification of 4-methyloctanoic acid with ethanol was studied at 35°C. Esterification in a batch reactor (molar ratio of 1:8 (acid:EtOH)) resulted in the isolation of the enantiomerically enriched product (ee p =81%) and substrate (ee s =93%). In order to integrate reaction and separation, liquid-vapour equilibria calculations were performed showing that an excess of ethanol results in a very low ester fraction in the vapour phase. Since this is undesirable for an integrated process of reaction and product removal, a repeated batch reaction was performed using a molar ratio of 10:1 (acid:EtOH). After six cycles (45% conversion) the ee of 4-methyloctanoic acid ethyl ester turned out to be 80%. For different E values the ee p was calculated for batch and repeated batch reactions. It was shown that in all cases the ee p was higher for the repeated batch reaction. However, the product is not enantiopure since the E value of the reaction is rather low at the low ethanol concentration used. An alternative approach would be the continuous separation of the product during the reaction. A mathematical model was developed to describe esterification in a packed bed reactor integrated with product separation. This model shows that integration of reaction and product removal in advance is not suitable either to obtain an enantiomerically pure product. Since the optimal reaction conditions (high ethanol concentration) and the optimal separation system (low ethanol concentration) do not match in this reaction, the preference is given to the batch reaction at high ethanol concentrations because in that case the highest enantioselectivity of the enzyme is obtained.     

L1 (LINE-1) retrotransposable elements provide a "fossil" record of the phylogenetic history of murid rodents

The single most difficult problem in phylogenetic analysis is deciding whether a shared taxonomic character is due to common ancestry or one that appeared independently due to convergence, parallelism, or reversion to an ancestral state. Mammalian L1 retrotransposons undergo periodic amplifications in which multiple copies of the elements are interspersed in the genome. Because these elements apparently are transmitted only by inheritance and are retained in the genome, a shared L1 amplification event can only be an inherited ancestral character. We propose that L1 amplification events can be an excellent tool for analyzing mammalian evolution and demonstrate here how we addressed several refractory problems in rodent systematics using L1 DNA as a taxonomic character.      

Dithiol- and NAD-dependent degradation of epoxyalkanes by Xanthobacter Py2

A broad range of epoxyalkanes was converted into the corresponding ketones by cell extracts of Xanthobacter Py2. Both 1,2- and 2,3-epoxyalkanes were degraded and in addition, the degradation of 2,3-epoxyalkanes in all cases was highly enantioselective. Conversion of a deuterium-labelled substrate indicated that the ketone product was probably formed indirectly via an hydroxy intermediate. Degradation of epoxyalkanes by Xanthobacter Py2 was dependent on both NAD and another, not yet identified, cofactor that was present in the low-molecular-mass fraction (LMF) of propene-grown cells. It is proposed that the LMF was involved in a reductive reaction step since it could be replaced by dithiothreitol (DTT) and various other dithiol compounds. Epoxyalkane-degrading activity was inhibited by the sulphhydryl blocking reagent N-ethylmaleimide (NEM). Inhibition by NEM and stimulation by LMF, DTT and other dithiols was effective only in the simultaneous presence of NAD.     

A proteomic approach identified growth hormone-dependent nutrition markers in children with idiopathic short stature

Background  

The broad range in growth observed in short prepubertal children receiving the same growth hormone (GH) dose is due to individual variation in GH responsiveness. This study used a pharmaco-proteomic approach in order to identify novel biomarkers that discriminate between short non-GH-deficient (GHD) children who show a good or poor growth response to GH treatment.