Theoretical and empirical population genetics
We are interested in theoretical and empirical population genetics. Our aim is to understand the role of natural selection and random processes in the evolution of plant species. We focus particularly on the consequences of plant-parasite coevolution and seed banking on genome evolution, using as a model system several closely related wild tomato species. We also work study the interactions between natural selection and genomic architecture in maize.
Background and rationale
Parasites have a negative impact on the fitness of their host, and can be responsible for drastic epidemics in human, animal or plant populations, with potentially large effects on biodiversity of ecosystems or on agro-ecosystems. With global trade and climate change, there is an increase of emergent invasive diseases such as viruses, bacteria or fungi driving host extinction, threatening hot spots of biodiversity (see the recent Fisher et al., Nature 2012), or affecting the sustainability of agro-ecosystems (e.g. Stuckenbrock and McDonald, Annu Rev Phytopathol 2008). Following the ‘Red Queen’ hypothesis (Van Valen, Evol Theor 1973), antagonistic interactions between host and parasite species are recognized as major fundamental forces driving rates of molecular evolution and rates of speciation (Paterson et al., Nature, 2010 ) and key factors organizing the Earth’s biodiversity (Emerson and Kolm, Nature 2005). Currently, the -omics era is bringing together two traditionally quite distinct disciplines in the study of host-parasite interactions: ecology and molecular biology. It is quickly becoming feasible to use ‘environmental genomics’ approaches to investigate the interactions between variable physical environments, host genetic variation and life-history traits.
Overview of research topics
Plant-parasite interactions provide particularly suitable model systems for molecular ecology studies as there has been a substantial emphasis in recent years on dissecting the genetic and molecular architecture of host-parasite specificity (Jones and Dangl, Nature 2006). Coevolutionary models suggest that reciprocal changes of allele frequencies over time at genes involved in host and parasite interactions can follow two simple scenarios (Woolhouse et al., Nature Genet 2002; Holub, Nat Rev Genet 2001): the ‘arms-race’ scenario, which is a series of recurrent selective sweeps, and the ‘trench warfare’ (or balancing selection) scenario, which maintains alleles at intermediate frequencies due to frequency-dependent selection. Using as a case study the gene-for-gene interactions (GFG) between host (plant or animal) resistance genes and parasite effectors, we have shown (with James KM Brown) that a fundamental mathematical condition, direct frequency-dependence selection, is necessary for the occurrence of balancing selection and is promoted by various plant and parasite life-history traits and ecological characteristics (review in Brown and Tellier Annu Rev Phytopathol 2011): e.g. seed banks and perenniality or parasite polycyclicity. The occurrence of arms race or trench warfare dynamics thus depends on a few key coevolutionary parameters (ecological and genetic costs). Moreover, environmental variability also has a crucial influence on coevolution (the ‘Geographic Mosaic Theory of Coevolution’: Thompson J.N., The Geographic Mosaic of Coevolution 2005), specifically in GFG interactions (Laine and Tellier, Oikos 2008). These theoretical developments lay the foundations for a quantitative understanding of the major ecological mechanisms driving the molecular evolution of host defence genes in natural populations.
A specific life-history characteristic of most plant species is to produce seeds, which may remain in the soil for long period of time (Fenner and Thompson, The Ecology of Seeds 2004). In general terms, the reproductive mode (seed or egg dormancy) is described as a bet-hedging strategy in plants (Evans and Dennehy, Q Rev Biol 2005), invertebrates (Daphnia: Decaestecker et al. Nature 2007, mosquitoes) and micro-organisms (Lennon and Jones, Nat Rev Microbiol 2011) to buffer against environmental variability. Bet-hedging is a strategy in which adults release their offspring into several different environments to maximize the chance that some will survive, thus magnifying the evolutionary effect of good years and dampening the effect of bad years. It also counter-acts habitat fragmentation by buffering against the extinction of small and isolated populations, a phenomenon known as “temporal rescue effect” (Honnay et al., Oikos 2008). Improving our understanding of seed bank evolution and its genetic underpinnings is thus important for the conservation of endangered plant species. However, little is yet known about the evolutionary and ecological mechanisms governing the evolution of seed dormancy. Seed banks also promote the storage of genetic diversity, increasing thus the effective population size (Kaj et al., J Appl Proba 2001) and decreasing among population differentiation (Vitalis et al., Am Nat 2005). We investigate the consequences of seed banks on coalescent trees and expected patterns of polymorphism (Tellier et al. PNAS 2011; Živković and Tellier 2012).
Wild tomato species are excellent model organisms for ‘molecular ecology’ studies with a focus on the effect of the spatial structure of populations (metapopulation). These species exhibit numerous patches of small sizes subject to extinction/recolonization, with migration among demes. They are found in South America in a great variety of habitats ranging from coastal plains, to the Chilean desert and to high altitudes of the Andes (above 3000 m)(Peralta et al., Syst Bot Monogr 2008; Nakazyto et al., Evolution 2008). S. peruvianum has a large geographic range, and is suggested to be a generalist species with respect to adaptation to abiotic stress. On the other hand, S. chilense has a smaller range distribution, is found in dryer and more extreme habitats (like the Atacama Desert). I have already shown that coalescent theory and new Bayesian methods are appropriate to study rates of adaptation in such complex ecological set-ups (Tellier et al., PNAS 2011; Tellier et al., Heredity 2011). These two species harbour different seed bank adaptations, with S. peruvianum having longer seed dormancy than S. chilense (Tellier et al., PNAS 2011). So far signatures of coevolution at the genomic level have been mainly studied in S. peruvianum (Hörger et al., PLoS Genet 2012, Rose et al., Mol Plant Pathol 2011). However, comparing these two species is of interest to test the molecular underpinnings of coevolution and/or adaptation to abiotic environments.
A major aim in plant and animal breeding is to find genes underlying key traits for agriculture (yield, disease resistance, meat quality, numbers of eggs,...).
More specifically, in project S3 of Synbreed, the Populations Genetics group has for fundamental objectives to develop new statistical methods to analyse genomic data from domesticated animal or crops. This amount to further develop the coalescent theory to take into account population structure, pedigree, domestication history in the analysis of next generation sequencing data.