Structural evidence for MADS‐box type I family expansion seen in new assemblies of Arabidopsis arenosa and A. lyrata

Abstract

SUMMARYArabidopsis thaliana diverged from A. arenosa and A. lyrata at least 6 million years ago. The three species differ by genome‐wide polymorphisms and morphological traits. The species are to a high degree reproductively isolated, but hybridization barriers are incomplete. A special type of hybridization barrier is based on the triploid endosperm of the seed, where embryo lethality is caused by endosperm failure to support the developing embryo. The MADS‐box type I family of transcription factors is specifically expressed in the endosperm and has been proposed to play a role in endosperm‐based hybridization barriers. The gene family is well known for its high evolutionary duplication rate, as well as being regulated by genomic imprinting. Here we address MADS‐box type I gene family evolution and the role of type I genes in the context of hybridization. Using two de‐novo assembled and annotated chromosome‐level genomes of A. arenosa and A. lyrata ssp. petraea we analyzed the MADS‐box type I gene family in Arabidopsis to predict orthologs, copy number, and structural genomic variation related to the type I loci. Our findings were compared to gene expression profiles sampled before and after the transition to endosperm cellularization in order to investigate the involvement of MADS‐box type I loci in endosperm‐based hybridization barriers. We observed substantial differences in type‐I expression in the endosperm of A. arenosa and A. lyrata ssp. petraea, suggesting a genetic cause for the endosperm‐based hybridization barrier between A. arenosa and A. lyrata ssp. petraea.