Population and speciation genomics

Following the generation of a draft sequence of the collared flycatcher genome (Ellegren et al 2012 Nature 491:756-760), we use Ficedula flycatchers as a model system for genomic studies of adaptation and speciation.

Population genetic processes affect how the genomes of structured populations differentiate and start to diverge. For example, gene flow been subdivided populations tend to homogenize genomes and thereby reduce differentiation. Moreover, local adaptation can enhance differentiation. In a speciation context, analyses of heterogeneity in genomic differentiation have sparked particular interest as it has been seen as a route towards identifying the genetic basis of speciation. Specifically, the frequent observation in several systems of distinct genomic regions with highly elevated differentiation (‘differentiation/divergence islands’) has been suggested to represent candidate loci for ‘speciation genes’.

However, there is increasing evidence that genomic regions with elevated differentiation can evolve by processes unrelated to speciation per se. The concept of linked selection, which is particularly pronounced in low-recombining regions, will locally reduce the effective population size and thereby enhance genetic drift of segregating variants. In turn, it results in elevated rates of differentiation in particular genomic regions without having to invoke adaptive processes. We use simulations and theoretical models to elucidate the role of selective sweeps and background selection to linked selection.

We have by now sequenced some 500 genomes of four recently diverged flycatcher species of the genus Ficedula. We study the evolution of the genomic landscape of differentiation to resolve the issues raised above (Burri et al 2015 Genome Res 25:1656-1665; Nater et al 2015 Syst Biol 64:1000-1017; Dutoit et al 2017 Proc R Soc Lond B 284: 20162756), and we use transcriptome data to address differentiation at the level of gene expression (Uebbing et al 2016 Mol Ecol 25:2015-2028). We use whole-genome sequences to infer demography (Nadachowska-Brzyska et al 2013 PLoS Genet 9:1003942; 2016 Mol Ecol 25:1058-1072), and we are increasingly focusing on functional aspects by analyzing the role of transposable element insertions (Suh et al 2017 Mol Ecol  doi: 10.1111/mec.14439) and by annotation of conserved non-coding elements using comparative genomics approaches (Craig et al 2017 Mol Ecol in press).

As recombination rate variation is an essential component in understanding both sequence diversity and divergence, we estimate recombination rates in flycatchers using linkage maps (Kawakami et al 2014 Mol Ecol 23:4035-4058), genome-wide data on linkage disequilibrium (Kawakami et al 2017 Mol Ecol 26:4158-4172) and direct detection of recombination events by pedigree sequencing (Smeds et al 2016 PLoS Genet 12:e1006044).

To the top

Molecular evolution

We seek to understand the contribution of different evolutionary forces – mutation, selection, recombination, demography and drift - to genome sequence evolution. We primarily focus our research on birds. They show a large heterogeneity in chromosome size, large variation in recombination rate across the genome, together with an evolutionary stable karyotype.

On-going molecular evolutionary studies in our laboratory follow several tracks. For example, we investigate the diverse consequences of recombination rate variation on genome evolution. We study the impact of Hill-Robertson Interference on gene sequence evolution, and investigate how GC-biased gene conversion affects the inference of selection in avian genomes (Bolívar et al 2016 Mol Biol Evol 33:216-227). We investigate the validity of the nearly neutral theory in avian species. Surprisingly, avian taxa seem to be an exception to this theory, since life history traits, commonly used as proxies for the effective population size, are not consistently correlated with measures of the efficacy of selection, such as the dN/dS ratio (Weber et al. 2014 Genome Biol 15:542). We investigate the apparent lack of correlation between life history traits and dN/dS in light of the pronounced signatures of GC-biased gene conversion in avian genomes.

We also perform theoretical studies and investigate the applicability of phylogenetic approaches to closely related species. Estimates of molecular evolutionary rates, such as dN/dS, are important measures for a wide range of evolutionary analyses. Its estimation is based on a phylogenetic approach, which rely on the indirect assumption that sequence divergence and species divergence are identical, an assumption which is generally violated and leads to a time-dependence of dN/dS (Mugal et al. 2014 Mol Biol Evol 31:212-231). We use Poisson random field models to derive analytical expressions of sequence divergence as a function of time and sample size (Kaj & Mugal 2016 Theor Pop Biol 111:51-64). This mathematical framework allows us to investigate if the joint usage of polymorphism and divergence data can improve the inference on the mode and strength of selection for closely related species.

In addition, we are interested in the evolution of regulatory sequences and to which degree genetic and epigenetic changes in regulatory sequences affect gene expression evolution. We develop methodological approaches to study gene expression evolution in natural populations (Wang et al. 2017, Genome Biol Evol 9:1266-1279). Using bisulfite sequencing, we investigate the impact of epigenetic changes on expression evolution.

To the top

Sex chromosome evolution and sex-biased gene expression

While males are the heterogametic sex in most animals with heteromorphic sex chromosomes, females are heterogametic sex in all birds (males ZZ, females ZW). This contrast is quite useful for tests of hypotheses concerning sex chromosome evolution and the evolutionary significance of sex-linkage.

Sex chromosomes have evolved independently numerous times in different lineages, yet they show many common features and thus represent a fascinating example of evolutionary convergence. The process of sex chromosome evolution has attracted considerable interest over the years and one important question has been to elucidate the evolutionary forces behind the cessation of recombination between proto-sex chromosomes, eventually leading to the formation of two quite discrete chromosome types. Other questions have focused on the genomic organization of sex chromosomes and the evolutionary implications of sex-linkage. This is related to sexually antagonistic genes and sex-biased gene expression.

Our work has since long been focused on avian sex chromosomes (Ellegren 1996 Proc R Soc Lond B 263:1635-1641; Fridolfsson et al 1998 Proc Natl Acad Sci USA 95:8147-8152; Berlin & Ellegren 2004 Proc Natl Acad Sci USA 101:15967-15969), and the formation of evolutionary strata in these (Handley et al 2004 Genetics 167:377-385; Nam & Ellegren 2008 Genetics 180:1131-1136; Yazdi & Ellegren 2014 Mol Biol Evol 31:1444-1453). We have recently sequenced and analysed the female-specific W chromosome of flycatchers (Smeds et al 2015 Nat Comm 6:7330), including the pseudoautosomal region (PAR; Smeds et al 2014 Nat Comm 5:5448).

A related line of research is the study of gene expression in autosomes vs. sex chromosomes, and in males vs. females. The finding that birds apparently lack wholesale dosage compensation of sex-linked genes (Ellegren et al 2007 BMC Biol 5:40) has inspired us and others to investigate the evolution of incomplete dosage compensation in a ZW system (Uebbing et al 2013 Genome Biol Evol 5:1555-1566; Adolfsson & Ellegren 2014 Mol Biol Evol 30:806-810). We use proteomic approaches to study dosage compensation also at the protein level (Uebbing et al 2015 Mol Biol Evol 32:2716-2725).

To the top

Conservation genomics

We analyze the genetic consequences of carnivore populations in Scandinavia being small and isolated, and develop tools for non-invasive genetic monitoring of these populations.

The Scandinavian wolf population is a striking example of the impact of isolation on genetic diversity and survival. Wolves were once extinct in Scandinavia, but in the early 1980s two individuals re-entered the peninsula and successfully reproduced. Cut off from the larger Finnish-Russian metapopulation by the Reindeer husbandry area to the North, the population remained very small (less than 10 individuals) for a decade. However, in 1991 a single male immigrant entered the population and reproduced, resulting in genetic rescue through an increase in both genetic diversity and population size (Vilá et al 2003 Proc R Soc Lond B 270:91-97). To this day the population remains characterized by prolonged periods of isolation with rare immigration events. We have made detailed investigations on how this has impacted on the levels and character of genetic variability (Ellegren et al 1996 Phil Trans R Soc Lond B 351:1661-1669; Hedrick et al 2001 Evolution 55:1256-1260; Flagstad et al 2003 Mol Ecol 12:869-880; Seddon et al 2005 Mol Ecol 14:503-511; Hagenblad et al 2009 Mol Ecol 18:1341-1351).

More recently, we have been investigating the genomic consequences this isolation. We are measuring the frequency and length of runs of homozygosity, distinctive genomic deserts devoid of genetic variation, that are signatures of inbreeding (Kardos et al 2017 Nat Ecol Evol doi: 10.1038/s41559-017-0375-4). In addition, we are applying statistical methods to reconstruct the original founders of the population, to estimate how much of their original genetic contribution persists within the population today. In combination, these approaches promise not only to provide a better understanding of the genomic consequences of population isolation, but also give vital information for the long-term management of this population of conservation concern.

In parallel and similar efforts, we have been studying the Scandinavian wolverine population. Here we provide service to Swedish and Norwegian management authorities by performing microsatellite and SNP genotyping of thousands of scat samples every year.

To the top