We have generated de novo reference assemblies based on long-read and short-read DNA and RNA sequencing for six Hordeum species (annual, perennial sister species) and, together with the PanHordeum Consortium, have generated annotated de novo assemblies for 22 Hordeum species (publication expected to end 2024).
Furthermore, we have used reduced representation sequencing of 80 accessions from 21 species, to reconstruct the Hordeum phylogeny. The phylogeny demonstrates that there were three independent transitions between annual and perennial life history in the Hordeum clade. Further, we used the phylogeny and genetic data to identify structural and genetic variations that consistently differentiate between annual and perennial species. This allowed the identification of candidate genes with roles in development and carbon allocation which we are currently testing.
We have trialed 52 annual and perennial accessions in outdoor experiments at the Botanic Garden of the University of Düsseldorf and scored 25 vegetative and reproductive traits over three consecutive seasons. This allowed us to identify traits and trait complexes that differentiate annual and perennial species. In particular, harvest index, allocation of N and C resources, and growth parameters differentiated between life history strategies. The dissection of life history transitions and meristem development in perennial species showed, that in contrast to many perennial species, perennial Hordeum species are not characterised by a juvenile phase or intrinsic seasonality. Seasonal flowering was only controlled by external cues, primarily day length and ambient temperature. In contrast to perennial rice and wheatgrass, perennial Hordeum species do not regrow from rhizomes, but from meristems at the crown of the plants.
Furthermore, we have generated interspecific crosses between annual H. intercedens and H. euclaston with perennial H. stenostachys and H. erectifolium. We have scored developmental and growth-related traits in the F1 plants and could show that perenniality is dominant in all these F1 crosses, whereas morphological and growth-related traits were intermediate. We have also conducted global transcriptome analyses in the parental and F1 plants to identify consistent cis- and trans-regulatory differences between the annual and perennial genomes. From the F1 plants, we have generated fertile BC1F1 and BC1F2 plants, which will be phenotyped during the next season in the Botanic Garden of the HHU.
Finally, we have started to generate Crispr/Cas transformants by either targeting the coding or regulatory regions for selected candidate genes. We have functionally analysed two candidate genes, which extended plant longevity by more than a year.
