Genomic consequences of shifts in mating system

Plants can reproduce in one of two ways; they can outcross, requiring pollen from another plant to fertilize their ovules, or they can self, where pollen from a plant fertilizes it's own ovules. 

Most of the Wright lab is interested in what happens on the genetic level when plants transition from being predominantly outcrossing to predominantly selfing. When these shifts happen to a population of plants there are two main changes that interest me. First, the effective population size of that population is decreased by about half of that of an outcrossing population of the same census size. Second, because of increased homozygosity in selfers linkage equilibrium decreases. The projects I describe below explore the consequences of these two changes on the genomes of outcrossers and selfers. 


Selection on non-coding regions in plant genomes


When the first genomes were sequenced we discovered many things, one of those was that a very large proportion of the genome of many organisms consists of DNA that does not 

code for protein. These non-coding regions make up ~98% of the human genome. We have learned that in some animals, like D. melanogaster, that both negative and positive selection can act very strongly. However, in other genomes, like humans, there seems to be less of a role for selection in non-coding DNA. 

However, to date the amount of selection acting on non-coding regions has not been quantified in plant genomes. Using Illumina sequence data from 13 Capsella grandiflora individuals I'm quantifying the amount of selection acting in both codign and non-coding regions genome-wide.

Collaborators: Emily JosephsStephen Wright

Imprinting in plant genomes

Most genes are expressed bi-parentally, that is from both the maternal and paternal copy. However, some genes are expressed exclusively from the paternal or maternal copy. This is well documented in mammalian genomes, often genes controling faster development (e.g. Igf2) are controlled by the paternal copy, while those that limit growth are controlled by the paternal copy (e.g. Igf2r). The kinship theory of genetic imprinting suggests that if multiple mating occurs, then fathers want to ensure their offspring recieve more resources, even at the expense of their half siblings. On the other hand, mothers want all of their offspring to do equally well, and want to equalize the resource distribution. Several hundered gened in humans, mice, and other mammals have been identified as imprinted. To date very few have been found in plant genomes. 

Most studies of imprinting in plants have been performed in the selfing species Arabidopsis thaliana. However, this may not be an ideal system to find imprinted genes. When plants self both the paternal and maternal genomes come from the same individual, and there should be no conflict in resource allocation among offspring. In outcrossing species, where individual plants can be fertilized by many pollen donors, conflict may be more common.

I am currently working to quantify the amount of imprinting genome-wide in the outcrossing species Capsella grandiflora. Using RNAseq of parent's and their progeny I will be assessing the imprinting status of all genes in the genome, at several different life stages. I will be following up this work by looking at the amount of imprinting in the selfing relative C. rubella, testing the hypothesis that conflict, and therefore the extent of imprinting, should be lowered in the selfer.

Collaborators: Young Wha Lee, Stephen Wright

Silencing of transposable elements and mating system

me_arvid.jpgTransposable elements (TEs) are mobile pieces of DNA with the ability to replicate themselves and insert into new locations of the genome, or transpose. They are one of the most common classes of gene on the planet, making up ~40% of the human genome and more than 80% of the genome of maize. Since most insertions by TEs are thought to be deleterious host genomes have evolved defenses against these elements in order to limit their activity. This defence network is mediated by small RNAs that target TE insertions for methylation, which then stops the TEs from being able to transpose, which we call silencing.

Shifts to selfing are associated with a decrease in the effective population size, Ne, which causes a decrease in the efficacy of selection. Because of this decrease we would expect to see an increased accumulation of TEs in selfing species, since selection cannot effectively remove them. However, this is not always the case.


My collaborator Arvid Ågren and myself have designed a simulation study to examine if loci that control the silencing of TEs can evolve faster, or to a more efficient level of silencing. Our preliminary results do indicate that silencing will evolve more quickly in selfers, despite the reduced Ne.

Collaborators: Arvid ÅgrenStephen Wright

Other Cool People

Greg Zynda