Abstracts of invited speakers
Departament de Genètica, Facultat de Biologia, i Institut de Recerca de la Biodiversitat (IRBio), Universitat de Barcelona, Spain
Chromosomal inversion polymorphism is widespread across the Drosophila genus, where in most cases it is considered to be adaptive. Despite the extensive knowledge accumulated on the geographical and temporal distribution of inversions in several Drosophila species, there is still limited information on the inversion generating mechanisms as well as on the effect of segregating inversions on nucleotide variation. Some advances on these issues as well as on the rates of chromosomal evolution across Muller’s elements will be presented.
Charles F. “Chip” Aquadro
Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY USA
Germline stem cells (GSCs) are capable of self-renewal but are also the progenitors of egg and sperm. We have found that a number of the genes that regulate maintenance and differentiation of GSCs show evidence of recent and/or recurrent natural selection. For example, a key “switch gene” for differentiation, bag of marbles (bam), shows strong positive selection for amino acid diversification that is strong and episodic across the genus Drosophila that results in 59 fixed amino acid differences between D. melanogaster and D. simulans (for a 433 amino acid protein). I will review our population genetic, functional, and comparative genomic analyses of these patterns of natural selection, including testing hypotheses as to the possible drivers of this selectively driven diversification. I will focus in particular on our efforts to test for a role of evolutionary conflict, particularly associated with the endosymbiont Wolbachia pipientis, in shaping natural selection at bag of marbles and other GSC genes in Drosophila.
Spencer C.H. Barrett
Department of Ecology and Evolutionary Biology, University of Toronto, Canada
Evolutionary transitions from outcrossing to predominant self-fertilization are frequent in hermaphroditic organisms, especially flowering plants where this change in mating system has occurred numerous times in unrelated families. Selfing is favoured in the short term because of the transmission advantage of selfing variants, and the ability of individuals to set seed without pollinators or mates at low density. However, predominant selfing has long been considered an ‘evolutionary dead end’ owing to the apparent ephemeral nature of many selfing lineages. Mating system transitions provide opportunities to investigate the demographic and genomic consequences of the evolution of selfing and the extent to which lower effective population size (Ne) reduces the efficacy of selection, potentially leading to the local extinction of populations. Here, I provide an overview of our research on the evolution of selfing in the annual, neotropical plant Eichhornia paniculata (Pontederiaceae) in which there is evidence of multiple independent origins of selfing from outcrossing, as a result of the destabilization of the floral polymorphism tristyly and spread of mating system modifier genes. Earlier studies indicate that interactions between stochastic forces and selection for reproductive assurance in small populations can explain the shift from outcrossing to selfing. Recent work has focused on investigating the genomic consequences of this mating system transition using range-wide sampling of populations and identification of nonsynonymous and synonymous polymorphisms segregating in outcrossing and selfing populations. Estimates of the distribution of fitness effects (DFE) and changes in the magnitude of selective coefficients (NeS) acting on mutations during the mating system transition provides evidence that the acquisition of selfing has been accompanied by genome-wide influences of reduced Ne and strong selection against deleterious recessive mutations, an example of purging of genetic load at the molecular level to accompany earlier evidence from experimental studies of fitness traits.
Institute of Evolutionary Biology, School of Biological Sciences, University of Edinburgh, UK
This paper attempts to combine the available information from applications of quantitative genetic and population genomic approaches to variation in natural populations of Drosophila. The aim is to estimate some important parameters of genetic variation in fitness, using a population genetics model of mutational effects on fitness components. The results suggest the existence of a class of mutations with much larger fitness effects than those inferred from sequence variability, and which contribute most of the standing variation in fitness within a population that is caused by the input of mildly deleterious mutations. However, deleterious mutations explain only part of this standing variation, and other processes such as balancing selection probably make a large contribution to genetic variation in fitness.
Institute of Evolutionary Biology, University of Edinburgh, UK
The defining feature of sex chromosomes is suppressed recombination. Recombination suppressed has evolved multiple times, both independently in different taxonomic groups of organisms, and at different times during the evolution of the sex chromosomes of animals and plants, forming “evolutionary strata”. Plants can give us fine-scale information about the progress of recombination suppression, because species with genetic sex-determination and sex chromosomes often have close relatives that are functionally hermaphroditic. The region of most interest for understanding what has led to recombination suppression is the pseudo-autosomal region (PAR), which still recombines, and it is hypothesized that maintenance of sexually antagonistic polymorphisms in PAR genes selects against recombination. We have been studying the PAR of a dioecious plant in the genus Silene, which as a heteromorphic XY sex chromosome pair. We have identified the linkage group that carries the fully sex-linked genes, and mapped it, in the dioecious species, and its homologue in a non-dioecious close relative. The PAR of the sex chromosome pair is largely homologous with two further linkage groups in the non-dioecious species, suggesting that these genes were recently added in large genome blocks onto the PAR of an ancestral XY pair. This region appears to be undergoing ongoing recombination suppression in the dioecious species. I examine results from estimating sequence diversity of PAR genes, to assess whether the region may include gene(s) with sexually antagonistic polymorphisms.
Andrew G. Clark
Department of Molecular Biology and Genetics, Cornell University, Ithaka, NY USA
Consideration of population structure may be informative about the processes regulating transposable element distribution and abundance. We had performed whole-genome sequencing of 84 lines of Drosophila melanogaster derived from Beijing, Ithaca (NY), Netherlands, Tasmania and Zimbabwe and from these sequences derived statistics on the distribution and abundance on all known TEs in melanogaster. We also developed libraries for short RNAs, and cataloged the distribution of piRNAs, in an effort to understand the host genome response to these elements. There was enormous variability in both the abundances of TEs and their respective piRNAs, including the relative “coverage” of each genome for its respective piRNAs. Consideration of some simple theoretical principles indicates that the system is in a state of considerable dynamic flux, and that fly genomes are constantly tuning their defenses to the vagaries of transposable element abundance.
Oscar E. Gaggiotti, Scottish Oceans Institute, University of St Andrews, UK
FST-based genome scan methods are frequently used to screen genome-wide patterns of DNA polymorphism consistent with signatures of positive selection. However, several evolutionary processes such as complex demographic histories, heterogeneous mutation rates across the genome, and purifying selection can increase the variance in FST leading to high false positive rates. In my talk I will present two different approaches that attempt to overcome this problem. The first one focuses on false positives due to a hierarchical population structure such as that observed for humans and other widely distributed species. The second one is more general and incorporates environmental data to estimate a locus x population interaction term to account for divergent selection; the excess variance in FST due to deviations from the underlying model are then captured by a locus-specific effect leading to a large reduction in the error rate.
Daniel L. Hartl
Department of Organismic & Evolutionary Biology, Harvard University
Cambridge, Massachusetts, USA
The Drosophila Y chromosome is a degenerated, heterochromatic chromosome with few functional genes. Nonetheless, natural variation on the Y chromosome in Drosophila melanogaster has substantial trans-acting effects on the regulation of X-linked and autosomal genes. Naturally occurring Y-linked regulatory variation (YRV) can be detected in somatic tissues and contributes to the epigenetic balance of heterochromatin and euchromatin at three distinct loci showing position-effect variegation (PEV). Moreover, naturally occurring polymorphic Y chromosomes differentially affect the expression of thousands of genes in XXY female genotypes in which Y-linked protein-coding genes are not transcribed. Interspecific divergence of Y chromosomes also contribute to male hybrid sterility between species. Using a series of Y-chromosome introgression lines in which the Y chromosome from either D. sechellia or D. simulans was introgressed into a common D. simulans genetic background, we find significant differences in expression for 122 genes, or 2.84% of all genes analyzed. Genes down-regulated in males with heterospecific Y chromosomes are significantly biased toward testis-specific expression patterns. These same lines show reduced fecundity and sperm competitive ability. Taken together, these results imply a significant role for Y/X and Y/autosome interactions in maintaining proper expression of male-specific genes, either directly or via indirect effects on male reproductive tissue development or function.
A meta-analysis of our full set of microarray studies indicates that YRV genes are more likely to be associated with the nuclear lamina than non-YRV genes and are also clustered in physical space in the nucleus (although not necessarily along chromosomes). In particular, YRV genes are preferentially localized to repressive chromatin in the BLUE and BLACK chromatin classes. These classes share a high affinity for binding of the proteins D1, LAM, and SuUR. Based on its DNA-binding properties and its affect on position-effect variegation of the w[m4] allele, we suggest that chromosomal protein D1 is a very strong candidate for mediating YRV.
Taken together, our results suggest that variation on the Y chromosome serves as a major source of epigenetic variation in natural populations and interacts with chromatin proteins and other regulators to modulate the expression of biologically relevant phenotypic variation.
Graduate University for Advanced Studies, Hayama, Japan
A pair of young duplicated genes occasionally have a complicated pattern of SNPs because they co-evolve through gene conversion. In such a case, the standard coalescent and diffusion theories are not suitable to apply. I here overview basic population genetics theory to understand the pattern of SNPs in duplicated genes where the effect of gene conversion taken into account. Two-locus diffusion theory tells what amount of polymorphism should be expected under neutrality, while the coalescent is useful for simulation-based analyses including neutrality test. Models with several modes of selection are also developed, which help to identify the footprint of selection in genome-wide polymorphism data. I will show some obvious signatures of selection from various species.
15 years of building on the work of Prof. Wolfgang Stephan –
insights on the distribution of fitness effects of new, segregating, and fixed mutations
Jeffrey D. Jensen
EPFL Lausanne, Switzerland
Understanding the relative frequencies and proportions of selective effects of newly arising, segregating, and fixed mutations - central to population genetics over the last century - stands as a major focus of the Jensen Lab. This underlying question has inspired our approach integrating toolsets from computer science, mathematics and statistics with evolutionary biology, medical and ecological genetics, experimental evolution, and population genomics. Despite drawing on this range of fields, at nearly every turn of our research the fundamental starting point for our thinking comes from the work of Prof. Stephan. I will here highlight how we have employed his important contributions into our current research studying adaptively important mutations as they arise, transit through the population, and ultimately achieve fixation – and I will discuss applications ranging from mouse to yeast to influenza virus. Finally, I will make the case that population genetics is on the verge of important breakthroughs in both our general understanding of the process of adaptation, but also in the ability of evolutionary analysis to provide important ecologically and clinically relevant insights to other research communities as well.
Department of Life Sciences and Division of EcoScience, Ewha Womans University, Seoul, Korea
Frequent adaptive evolution of influenza virus is not only a major global public health concern but also of great interest in evolutionary biology, as the abundance of longitudinal (time-series) viral sequences samples provides unique opportunity to observe adaptive evolution occurring in real time and to test predictions of classical and novel theories in population genetics, including selective sweeps and background selection. We examined the patterns of sequence variation in the population of seasonal influenza virus H3N2. Adaptive evolution is believed to occur through frequent fixations of advantageous variants mostly due to selective pressure to escape host immunity by changing the antigenic structure of viral proteins. We examined the temporal patterns of allelic substitutions that accumulated over the last 40 years. Our analysis based on computer simulations and Watterson’s classical theory indicates that basic genealogical process, together with the unique viral feature of high mutation rate over non-recombining segments, can generate clustered fixations of multiple mutant alleles, which were previously interpreted as the result of selective sweeps. However, the presence of recurrent positive selection is also required for generating the qualitative profile of clustered frequency trajectories that match actual data. This result implies that both adaptive and non-adaptive processes are importantly contributing to the evolution of H3N2 sequences. Indeed, we find that recurrent selective sweeps alone may not be responsible for the extremely low effective population size of global H3N2 population. Skewed distribution of viral progenies, background selection, and complex metapopulation dynamics, each of which can significantly reduce the level of variation, are likely to combine to determine the effective population size of H3N2.
Charles H. Langley
College of Biological Sciences, UC Davis, USA
The DNA double helices of the eukaryotic genome are highly compacted and functionally organized via their intimate associations with DNA-binding proteins. While known scales and complexities of this chromatin organization and its functional interactions are advancing, the elemental structure, the nucleosome has long been recognized. 80% of base pairs of a eukaryotic genome are in 147 bp segments wrapped around a highly conserved histone octamer core (a dimer of four distinct histones). Though the genomic positions of nucleosomes exhibit characteristic patterns over cells in tissues, their mobility along the genome and their modifications have been associated with virtually all genomic processes, including gene expression, replication, mutation, recombination, development and evolution. Furthermore, conserved (albeit relatively low energy) molecular interactions between characteristic nucleotides and specific histone moieties have been widely observed and interpreted as essential to conserved structure and many known functions.
From the perspective of evolutionary genomic variation the DNA sequences of individual nucleosomes can be examined on these two scales. Firstly, variation in the periodicities of polymorphism and divergence can be associated with the diversity of annotated function, including the patterns of nucleosome positions. Secondly, divergence among species in these patterns of nucleosome positioning can be characterized and interpreted both in terms of the local genomic function, as well as associated lineage-specific changes in sequence.
Recently published and an ongoing study of the nucleosomal sequences from Drosophila melanogaster, simulans and yakuba embryos show a clear ≈10 bp periodicity in divergence and polymorphism. The frequency spectra of these polymorphisms are consistent with systematic selection (or biased gene conversion) for the characteristically “preferred” mutation which often have straightforward structural interpretations in the canonical nucleosome. An initial examination of lineage specific divergences in nucleosomal position will be presented and discussed.
CAS-MPG Partner Institute for Computational Biology (PICB), Shanghai Institutes for Biological Sciences (SIBS), Chinese Academy of Sciences (CAS) Shanghai, P.R.China
A long-standing interest in evolutionary biology is to study the adaptive evolution driven by positive selection. About 12 years ago, I followed the selective footprint of Wolfgang Stephan and studied on the adaptive evolution of Drosophila. Together with Wolfgang, we proposed to use the genome-wide joint site frequency spectrum (JSFS) and a coalescent-based likelihood function to estimate the demography of Drosophila. Given the estimated demography, we detected genes positively selected in Drosophila. It has been further demonstrated that this strategy is a very reliable approach to map candidate genes for positive selection, and the JSFS-based method has also been popularly used to analyze human population data. After I moved to Cologne and then back to Shanghai, I kept working on this subject. Recently, together with Thomas Wiehe, we analyzed the statistical features of unbalanced tree in the neutral model and further proposed a new neutrality test that integrates the information of tree topology and the branch lengths of unbalanced tree. Simulations based on wide-range parameters demonstrated that the new test is robust to different demographic models and has a high statistical power to detect recent positive selection. Especially, this test does not need the information of outgroup, and thus will not be affected by the mis-inference of derived and ancestral variants of segregating sites due to multiple hits. This work provides a reliable method that can distinguish selection from demography when single-locus or genome-wide DNA polymorphism data is available.
Department of Biology, University of Rochester, USA
Adaptation is now the focus of considerable empirical and theoretical work in population genetics. I review some of the theory of adaptation. This work falls into two classes:
i) the genetics of adaptation in stable populations (e.g., Fisher’s geometric model and Gillespie’s mutational landscape model); and
ii) the genetics of adaptation in populations threatened with extinction.
I focus on this second class of problem, emphasizing recent theoretical progress in understanding the population genetics of “evolutionary rescue.”
Institute of Molecular Biology and Biotechnology, Foundation for Research and Technology-Hellas, Heraklion, Crete, Greece
Site-Frequency-Spectrum (SFS) based methods are powerful and adequately fast to detect selective sweeps in full genome data. Thus, they have been implemented, in various flavors, in modern neutrality tests, for example SweeD or SweepFinder. In brief, a log-likelihood value for a selective sweep model at a certain genomic location, is calculated for each single nucleotide polymorphism in the proximity of the selective sweep. Log-likelihoods are then summed up to assess the total likelihood of a sweep at a certain location. Thus, current methods use each SNP as an independent piece of information, neglecting information that potentially springs from combinations of them. Moreover, to calculate the likelihood value for a sweep they do not take into account demographic changes that the population may have experienced in the past. Here, we propose a simple simulation-based approach to
(i) include information from the demographic model in the calculation of the sweep likelihood and
(ii) to use information from pairs of SNPs instead of single SNPs.
Preliminary results show that this approach outperforms modern SFS-based methods.
Department of Biology, University of Rochester, USA
Sex chromosomes play a special role in the evolution of postzygotic reproductive isolation. For reasons that remain unclear, the X chromosome accumulates genetic factors that cause interspecific hybrid sterility several times faster than the autosomes. The rapid evolution of hybrid incompatibility on the X almost certainly contributes to Haldane’s rule, the large X-effect, and the relative paucity of X-linked interspecific gene flow between species.
I will present findings from our efforts to determine why the X is a hotspot for hybrid incompatibilities in Drosophila. Using a ultrahigh-resolution genetic mapping approach, we seek to determine the molecular functions and evolutionary histories for a panel of X-linked hybrid sterility genes. We also test several general explanations for the special role of the X chromosome in hybrid male sterility and make new discoveries about the unusual transcriptional regulation of the X in the Drosophila male germline.
Genetics and Physiology Unit, University of Oulu, Finland
Many forest tree populations show steep latitudinal clines in quantitative traits related to adaptation, but very little genetic differentiation with respect to random markers in the genome, showing that there likely is extensive gene flow between populations. We examine what such a cline. We have quantified patterns differentiation in timing of bud set of seedlings. This differentiation, based on photoperiodic reactions has likely arisen quite recently. Based on analyzing long term common garden experiments, stabilizing selection at individual sites is very strong, especially in the northern areas. Candidate gene based association studies in populations in common garden studies show small effects of alleles at individual loci. Overall, the genomic levels of differentiation are very limited (based on SNP and exome sequencing data), but shallow clines are found at some candidate loci. No loci show steep clines predicted by some model.
Yun S. Song
University of California, Berkeley, USA
Numerous empirical studies in population genetics have employed the sample frequency spectrum (SFS), which provides a highly efficient dimensional reduction of large-scale population genomic variation data into just a few numbers. Despite their popularity, the accuracy of inference methods based on the SFS has remained challenging to characterize analytically. In this talk, I will present some recent theoretical results on this problem.
Max-Planck Institute for Evolutionary Biology, Plön, Germany
Populations and closely related species can be genetically and phenotypically distinct, yet not fully reproductively isolated. This can lead to genetic exchange upon secondary contact. Such exchange has traditionally been considered as a problem for the respective populations and species, since it could result in a disruption of co-adapted gene complexes. However, it is becoming increasingly clear that hybridization and introgression of foreign genetic material can also have beneficial effects for the respective populations. I will present our most recent results that we have obtained from large scale studies of natural populations of the house mouse. These results suggest that adaptive introgression can also occur across large geographical distances, i.e. is not restricted to direct contact of the populations. In the case of the mice, it seems possible that only a few animals accidentally transported by humans into other areas can trigger the introgression of haplotypes that subsequently spread in the affected population. This suggests that the genetic variation that is built up under allopatric conditions is not necessarily maladaptive in a different genomic context, but could also contribute to new adaptations.
University of Frankfurt, Germany
Suppose a highly beneficial mutant appears in a geographically structured population and eventually will go to fixation. What can one say about the time to fixation, depending on the migration rates and the selection coefficient? This question was posed several years ago by Wolfgang Stephan. The process of its answering has led to nice mathematical insights, and to a fruitful collaboration within a geographically structured subpopulation of German probabilists. The type frequencies in the d colonies constitute an interacting diffusion in the d-dimensional unit cube which starts in (0,...,0) and eventually reaches (1,...,1). As it turns out, the entrance laws from (0,...,0) of this diffusion have a beautiful "Kingman paintbox" representation in terms of two mutually coupled ancestral Ancestral Selection Graphs, with a nice interpretation in terms of "potential ancestral lineages".
In my talk I will explain these concepts, a bit of their mathematics and the "biological ideas" behind them, and will describe how all this allows to analyze the asymptotics of the time to fixation, and to distinguish various parameter regimes.