ICTS

In a simple, constant environment does evolution continue forever? Does extensive diversification via small genetic and ecological differences? What are general evolutionary consequences of organismic complexity? Hints from long term laboratory evolution experiments and from genomic data of within-species bacterial diversity motivate considering these questions. Several simple models of evolution with small ecological feedback will be introduced, with the high dimensionality of phenotype space enabling mathematical analysis.

Populations often adapt to spatially heterogeneous environments via substitutions or allele frequency changes at many loci spread across the genome. Subsequent contact between diverged populations can produce unfit hybrids, which inhibits genetic exchange across the genome, thus setting the stage for further divergence and speciation. How such polygenic divergence builds up or is maintained depends on the coupled dynamics of multiple loci under selection, making it challenging to predict. In this talk, I will outline simple theoretical approximations based on effective migration rates that capture how coupling or linkage disequilibria (LD) between loci influences the maintenance of polygenic adaptive divergence. This analysis allow us to clarify how genetic architecture (i.e., the numbers, selective and dominance effects and linkage relationships of selected variants) influences divergence in the face of gene flow and genetic drift. I will conclude my talk with a broader perspective on the unreasonable effectiveness of effective parameters in population genetics.

For over a century, most biologists have been convinced that all aspects of biodiversity have been driven entirely by natural selection, with stochastic forces and mutation bias playing a minimal role. However, this is not the case at the molecular and cellular levels, where diverse traits scale with cell/organism size in ways that cannot be explained by optimization and/or speed vs. efficiency arguments. These include aspects of gene/genome architecture, intracellular error rates, the multimeric nature of proteins, swimming efficiencies, and maximum growth rates.

Although natural selection may be the most powerful force in the biological world, it is not all powerful, and the power of random genetic drift ultimately dictates what selection can and cannot accomplish. Many prokaryotes may reside in population-genetic environments where the limits to selection are indeed dictated only by the constraints of cell biology. However, in the eukaryotic domain, larger organism size is typically associated with a reduction in effective population size (Ne), enabling the accumulation of very mildly deleterious mutations, which in turn induces coevolutionary side effects leading to more complex and less efficient phenotypes.

This general conclusion is embodied in the drift-barrier hypothesis, which postulates that traits under persistent directional selection become stalled when further increments in improvement are thwarted by the power of random genetic drift. Integration of biology’s three engines of quantitative theory – population genetics, biophysics, and biochemistry, combined with observations from cellular bioenergetics, is providing a platform for the emergence of a formal field of evolutionary cell biology.

The interplay between natural selection and migration is predicted to shape the architecture of adaptation in different ways, depending on whether the direction of selection is spatially homogenous or heterogeneous. When different populations experience selection towards a similar phenotypic optimum, there is no tension with migration and "global adaptation" proceeds in manner similar to that predicted for a single population. By contrast, when populations are selected towards different optima, "local adaptation" occurs, which tends to favour architectures driven by fewer, larger, and more tightly clustered alleles than global adaptation. Despite this clear theoretical prediction, there have been few, if any, comprehensive tests in natural populations. Here, we bring together genome sequence data from thousands of individuals from 25 species of plants to compare signatures of repeated selective sweeps (global adaptation) with those of genotype-environment association (local adaptation). We deploy a common bioinformatic pipeline to call variants in all datasets, reconstruct orthology relationships, and test for repeated signatures across species. In both the local and global adaptation analyses, we find a large number of genes with evidence for repeated signatures across multiple species, including many genes with previously established functions in response to biotic and abiotic stress. We then use RNAseq datasets to build gene co-expression networks, which provide an estimation of the pleiotropy of each gene (more connections/centrality == more pleiotropy). As pleiotropy is generally positively related to allele effect size, we would expect more pleiotropy for local adaptation and less pleiotropy for global adaptation due to the interplay between migration and selection. Our results are remarkably consistent with this prediction, with significant enrichment for high pleiotropy among the genes driving local adaptation, and low pleiotropy among the genes driving global adaptation.

Most adaptive traits are polygenic with many underlying loci. The genetic architecture of these traits specifies how the phenotypes can be changed by mutations, but many factors determine which of these loci will be used during adaptation. Additionally, adaptation of complex traits in replicate populations with phenotypic convergence can occur through selection of different sets of loci. The analysis of time-series phenotypic and genomic data in replicates of evolving populations is crucial for understanding the adaptive architecture of a trait. I will demonstrate how experimental evolution emerges as an exceptional approach for generating precisely this type of data. Drawing on empirical evidence, I will show that temporal phenotypic data enable the identification of adaptive patterns, even when the selected trait(s) are unidentified, as is often the case in natural and experimental populations. Finally, I will discuss the next generation of evolution experiments, designed to mimic the assumptions of theoretical models of polygenic adaptation. These experiments help in testing the predictions of theoretical models and exploring the effect of factors such as selection strength on adaptive responses.

Ecosystems are threatened worldwide by anthropogenic degradation, fragmentation and rapid climate change. Plants are critical for nearly all food webs and thus for the functioning of ecosystems. Many plant species are going extinct, at least on some part of their range. So in the future, to restore complex ecosystem and protect biodiversity, we must understand the genetic resources of endangered plant species. Arabidopsis lyrata ssp. petraea is a close relative of the model species Arabidopsis thaliana. In some part of its range, populations are becoming increasingly scarce. I will report work in my lab that examines three crucial aspects of genomic resources: the amount of deleterious variation, the fraction of additive genetic variance in populations and the optimization of complex polygenic traits. In doing so, I will illustrate how useful the study of gene expression variation is.

I present shared work with Benjamin Wölfl and Ilse Höllinger on the adaptive architecture of a quantitative trait that with a genetic basis of moderate size (oligogenic adaptation).

Multiple comparisons, combining p values, meta-analysis, model selection