UC Berkeley

Polyploidy, the condition of having more than two sets of chromosomes, is very common in the plant kingdom. However, the role of polyploidy, or whole-genome duplication (WGD), in plant evolution is far from clear. In some cases, polyploidy appears to be an engine of saltational evolution, while in other cases it seems to have little appreciable effect. To understand the role of polyploidy in plant evolution, we must look to the genetics and genomics of the WGD event itself. The progenitors’ genomes play a crucial role in establishing the polyploid’s genetic stability and ultimate evolutionary trajectory. We have investigated genome structure and evolution in Brachypodium hybridum (2n=4x=30). The purple false brome B. distachyon (2n=2x=10) is a well-established model organism, and it is also one of the progenitors of B. hybridum, the other being B. stacei (2n=2x=20). Despite the availability of many genetic resources for B. distachyon, little is known about the allotetraploid B. hybridum. Here, we describe several genomic features of two independent natural lineages of B. hybridum. In chapter 1, I describe the transcriptional landscape of the model B. hybridum accession ABR113. We found that polyploidy had no appreciable effect on the transcriptome, and differences between progenitor and subgenome were within the level of variation we would expect between accessions of the same diploid species. We used a novel and straightforward analytical approach to cross-species RNA-seq that may be valuable to other researchers. We also demonstrated the importance of including the progenitors in studies of polyploid gene expression, as ‘parental legacy’ appeared to be the main driver of gene expression. This was part of a larger effort to survey the genome of ABR113, which was published in Gordon et al. (2020). Chapter 2 constitutes the main project of my doctoral research. This project describes the genome of B. hybridum accession Bhyb26. Bhyb26 is an older lineage of B. hybridum: the Bhyb26 WGD event is estimated to have occurred 1.4 million years ago, while the ABR113 WGD occurred approximately 140,000 years ago (Gordon et al. 2020). Both lineages possess the expected karyotype, that is, a D subgenome of 10 chromosomes and an S subgenome of 20 chromosomes, with no major rearrangements or sequence loss. However, we found that the Bhyb26 genome shows subtle genomic changes consistent with relaxed natural selection where the younger ABR113 genome did not. When we searched for biased genome evolution favoring one subgenome in Bhyb26, the results were mixed. We conclude that Bhyb26 is evolving toward the diploid state in a very gradual and largely unbiased manner. This study constitutes a rare glimpse of diploidization “caught in the act” on two evolutionary timescales. Chapter 3 is a survey of the 3D genome topology of B. hybridum ABR113 and its progenitors using Hi-C technology. During the course of our survey, we stumbled onto a mysterious 3D chromatin structure in the B. hybridum genome. We find evidence that this 3D chromatin structure, composed of mutually contacting loci akin to the KEEs/IHIs described in Arabidopsis (Feng et al. 2014; Grob et al. 2014), was passed down from progenitors to polyploid. The two separate chromatin interaction networks merge to form a single network in the polyploid, which we refer to as the KNOT, in keeping with Grob et al. (2014). Transposable elements (TEs) in the Brachypodium KNOT bear certain characteristics that are consistent with proposed functions of the Arabidopsis KNOT in TE defense, though the participating loci are not as TE-rich as those hypotheses predict. This study approaches the question of ‘parental legacy’ from an unusual angle: the inheritance of 3D chromatin structural features. In summary, we find that B. hybridum belongs in a class of polyploids that show a subtle and unbiased, rather than a disruptive, genomic response to WGD. We provide theoretical arguments explaining this observation. B. hybridum supports the hypothesis that even an allopolyploid from a very wide cross can undergo gradual, unbiased evolution provided that the progenitor genomes are similar in their TE load.