Skip to main content

Mating system and the evolutionary dynamics of hybrid zones

Final Report Summary - PLANT HYBRIDIZATION (Mating system and the evolutionary dynamics of hybrid zones)

Final publishable summary report

My Marie Curie grant resulted in four research projects: (i) Evolutionary pathways for the generation of new self-incompatibility haplotypes in the SRNase/F-box non-self recognition system (ii) Does self-incompatibility facilitate introgression? (iii) Development of S allele specific primers from RNASeq data, and (iv) A comparison of the genomic signals of balancing vs. divergent selection in snapdragons. Project (i) and (ii) were developed in the final year when I realized there were going to be delays in finalizing the bioinformatics for the S allele primers and in the release of the Antirrhinum genome paper (which is required before I can publish (iv)). Progress for each of these projects is outlined below.
(i) Evolutionary pathways for the generation of new self-incompatibility (SI) haplotypes in the SRNase/F-box non-self recognition system. In this project I worked with three other post-docs (Dr. Tadeas Priklopil, Dr. Katka Bodova and Dr. David Field) to develop an analytical model and stochastic simulations to examine the conditions under which new self-incompatibility alleles evolve in a non-self recognition SI system, such as in Antirrhinum (Snapdragons). Despite theory on novel S allele evolution in self-recognition systems, there are currently no models to assess this in non-self recognition systems. Consequently, this research fills an important knowledge gap and will make an important contribution to the research field of plant mating system evolution. In our paper we show that novel S haplotypes can evolve across a range of parameter values (inbreeding depression and self-pollination) but that this varied with evolutionary pathway. Moreover, S allele diversification was observed through pathways that maintain SI as well as through self-compatible intermediates. We found that the parameter space for diversification was larger in a finite compared to infinite population. However, in finite populations, the number of haplotypes observed under conditions of high inbreeding and moderate to high self-pollination was less than that commonly observed in natural populations. Higher mutation rate and number of potential haplotypes resulted in shorter lifespans and a greater turnover of SI classes, with evolution though the SI pathway more common for S-haplotypes with long life span and high frequency. This project is in the final stages of manuscript preparation and will be submitted to Genetics in November 2017.
(ii) Does self-incompatibility facilitate introgression? The modeling in this project was undertaken in July 2017 with a summer intern at IST Austria, Tanmay Dixit, an undergraduate from Cambridge University, UK. This grant therefore provided research training and experience for undergraduates considering doctoral studies and a career in research. Our results suggest that under conditions associated with strong negative frequency dependent selection (low diversity and high differentiation in S alleles), self-incompatibility can facilitate gene flow and overcome the effects of even strong selection against hybrids. The higher effective migration rate under these conditions could therefore facilitate introgression and allow the persistence of hybrid zones, even in the presence of barriers to gene flow. We found that this tension between different parts of the genome is also dependent on the underlying genetic architecture of the selected loci. Here the degree of linkage between selected loci strengthened the effect of hitchhiking on patterns of introgression, so that low recombination between loci had two contrasting outcomes; increasing effective migration rate under strong negative frequency dependent selection, and decreasing it with strong selection against hybrids. This implies that the genomic architecture of loci involved in both self-incompatibility and reproductive isolation will determine patterns of gene exchange in hybrid zones.
(iii) Development of S allele specific primers from RNASeq data. Some of the initial population genetics questions outlined in my Marie Curie proposal required the development of S allele specific primers. We have, however, had difficulties in developing these markers. Despite this, we are still working on marker development and hope to have these finalized by the end of 2017; as such, once finished, this will provide a new method for developing primers for other self-incompatible species with RNAse based self-incompatibility systems. One problem that has occurred during the project is the delay in the finalizing and release of details of the S-locus in the Antirrhinum reference genome from our collaborators in China (lab of Prof. Yongbiao Xue). This lack of information has provided challenges for the bioinformatics part of this project and we are currently exploring denovo assembly methods as an alternative solution. I also hope to undertake some long-read sequencing using the Oxford Nanopore systems to help generate sequence that will span the SRNase genomic region, facilitating primer design.
(iv) A comparison of the genomic signals of balancing vs. divergent selection in snapdragons. As outlined in my project proposal, systems such as Antirrhinum, with well-characterized flower colour genes (divergent selection) and known S-locus (balancing selection) provide the opportunity to examine the role of adaptive introgression in hybrid dynamics and if patterns of introgression vary across the genome for loci subject to different selection pressures. As such I aimed to compare genomic signals of divergent and balancing selection to the genomic background for both pooled and individual whole genome data for two sub-species of Antirrhinum majus (A. m. striatum and A. m. pseudomajus). Specifically, I used both pooled (50 individuals) and individual whole genome data (five individuals). Results are in line with our expectations of genomic signals of these different forms of selection. For example, for the S-locus genomic regions under balancing selection (outlined in Figure 1, attached), we expect high genetic diversity and low differentiation. Moreover, at the self-incompatibility locus, we may expect a higher frequency of rare alleles and allele frequency distributions that are skewed towards low frequency alleles compared with other regions of the genome. I found that Site frequency spectrum data demonstrates a high frequency of rare alleles for both A. m. pseudomajus and A. m. striatum. This skew towards low frequency alleles is combined with a deficit of moderate frequency and common alleles within the F-box genes, as expected under balancing selection. This aligns with our expectations and provides an interesting dataset for publication. The reason that this has not been published yet is that it uses the sequences for the 36 SLF genes provided by Prof. Yongbiao Xue. Once these sequences have been published as part of their S-locus genome paper, I will be able to finalise and publish this MS.

Related documents