The molecular basis of postzygotic hybridization barriers in plants

Different species are often genetically isolated from one another by a 'reproductive barrier' that prevents them from having common offspring. In this project, we study barriers that prevent the embryo from developing normally when gametes from two species merge. In plants, the embryo develops inside the seed, which in many plants additionally contains nutritive tissue (endosperm) that the new plant needs to grow. We are studying embryo and endosperm development in seeds resulting from crosses between different species (hybrids). We further investigate which genes are responsible and to what extent the result is affected mostly by the mother or father. We are working with two genera in the mustard family (Brassicaceae): Arabidopsis (Thale cresses) and Draba (Whitlow grasses). Results from Arabidopsis show that there is a strong barrier (with none or few surviving seeds) when crossing two diploid species, A. arenosa (Sand rock-cress) and A. lyrata (Northern rock-cress). We found that this is owing to failure in the formation of endosperm depending on which species is mother in the cross. If A. lyrata is mother, no cellularisation takes place (meaning that cell walls are not formed) but if A. arenosa is mother cellularisation is much too fast, in both cases with a fatal result. In nature, in addition to diploid populations, there are also polyploid populations of both species. Polyploid plants have a doubled genome (a tetraploid thus has four chromosome sets instead of two as in a diploid). One of our hypotheses was that genome duplication (polyploidy) may contribute to break down the barrier between species, and this is exactly what we found. When tetraploid A. lyrata is involved in crossings with diploid A. arenosa, endosperm development is normal and viable seeds are produced. However, the same does not happen when tetraploid A. arenosa is involved in the cross. In the genus Draba, we cross three species (D. nivalis, D. fladnizensis, D. subcapitata). The crossings that we have done indicate that a similar endosperm barrier as we found in Arabidopsis does not exist between Draba species, and that other reproductive barriers may be working. To investigate which genes are involved in the barrier between A. lyrata and A. arenosa, we have DNA sequenced the whole genome of the two species, to have these as resources when interpreting our results. We found that the genome of A. lyrata is somewhat larger (ca 201 mill. Base pairs) than the genome of A. arenosa (ca 179 mill. base pairs). We have further sequenced transcribed genes in young seeds from parental and hybrid plants. We use these data to examine gene expression before and after endosperm cellularisation to see if the expression of specific genes is connected with survival of the offspring, and if it matters whether it is the contribution (allele) from the mother or father that is expressed (genetic imprinting). We use bioinformatics analyses to check out all transcribed genes that show specific expression in the parents and may be important for the barrier. We have started analysing a group of MADS-box type-1 genes known to be involved in endosperm development. We characterise and compare the MADS-box gene family with what we find in other genome-sequenced species in the genus (and the Mustard family). So far, we found that MADS-box gene copies in A. arenosa and A. lyrata are quite similar and we have specifically for one gene (AGL36) shown that it is only expressed if inherited from the mother. This is also seen in hybrids from crosses between the two. We show, however, that the paternally inherited allele is reactivated in hybrids between A. arenosa and A. thaliana (Thale Cress, another close relative). This suggests that the gene is important for the endosperm barrier, but the result is dependent on which species are being crossed. As part of these experiments, we also found that both the genotype of the parental species and temperature could affect the barrier and thus the result of crossing between two species. Only a slight reduction in temperature during seed development increases the survival of the hybrid seeds. If this turns out to be a general trend, and the also the case in hybridisation between other species, this may contribute to a better understanding of the reproductive barrier between two species. In addition, it is easy to imagine that such a temperature effect can lead to several cases of hybridization under changing climatic conditions, which in turn can result in two species becoming genetically identical over time and no longer considered different species, or alternatively (especially if coupled with genome doubling) that new hybrid species are formed. A better understanding of how reproductive barriers between species can be strengthened or weakened may also have transfer value to breeding of agricultural plants.

The project has improved our common knowledge on one of the core issues in evolution, the underlying causes of the evolution of reproduction barriers, which are driving forces for speciation and maintaining biodiversity in the long term. Our results deliver new knowledge for understanding the first reproductive barrier after fertilization in plants. The indication that hybridization may be affected by environmental factors, such as temperature, is especially relevant in a scenario of climate change as we are experiencing now, especially if it turns out to be a general trend for endosperm-based barriers in several plant groups. Increased knowledge on this subject may also pave the ground for designing genetic strategies to overcome hybrid barriers, holding great potential for plant breeding, especially if our results on how to strengthen or weaken hybrid barriers can be transferred to agricultural important plant species.

The evolution of reproductive isolation is a classic problem in evolutionary biology. Plant species are typically isolated by different pre- and postzygotic barriers in potentially complex interactions. Much progress has been made in characterizing individual components of reproductive isolation but still the evolutionary forces responsible for development of postzygotic barriers are virtually unknown. The aim of the project is to elucidate the molecular basis of postzygotic hybridization barriers. Our approach is unique as it integrates detailed observation of embryo and endosperm development in hybrid seeds with identification of genes and mechanisms responsible for the barrier. To test our hypotheses, we will cross species within two Brassicaceae genera (Arabidopsis and Draba), look for similar patterns of aberrant seed development and relate these to underlying imprinting programs and gene networks. As part of the experimental setup, we will 1) characterize seed developmental phenotypes of postzygotic barriers in crosses between different species, different ploidy levels and cryptic species; 2) characterize maternal and paternal expression patterns in hybrid seeds using mRNA-seq; 3) determine imprinting status of selected candidate genes; and 4) integrate the results to improve our common knowledge on one of the core issues in evolution - the underlying causes of the evolution of reproduction barriers, and thus speciation. In addition, increased knowledge on this subject will pave the ground for designing genetic strategies to overcome barriers which may hold great potential for plant breeding.