Study of the effect of DNA-methylation on meiotic recombination in Arabidopsis thaliana and barley (Hordeum vulgare)

Meiotic recombination underpins both applied plant breeding and gene mapping in fundamental research. However, in large-genome crops such as cereals (including barley; Hordeum vulgare), recombination generally occurs close to the telomeres with around 30% of the genes rarely, if ever, recombining. Understanding recombination in cereals is therefore crucial. So far, most meiotic research in plants has focused on understanding mechanistic aspects of the formation of crossovers (CO), but little has been centred on the effect of epigenetic markers, including DNA-methylation, on recombination. However, manipulating the epigenome could have the potential to release novel combinations of genetic diversity. Maintenance of pre-existing DNA-methylation is mainly driven by Methyltransferase 1 (Met1), which is involved in CG-methylation. In Arabidopsis thaliana, when compared to wild type (WT), hypomethylated met1 mutants exhibit a higher CO frequency at the ends of chromosomes and decreased levels in the peri-centromeric regions. However, the overall number of COs in the genome remains constant. Similar trends were also observed in DNA-demethylation 1 (ddm1) mutants in Arabidopsis. In this study, we aim to manipulate DNA-methylation in barley and Arabidopsis, both transiently and by mutagenesis. Zebularine, a demethylating cytidine analogue, was applied to Arabidopsis and barley F1 hybrid seeds in an attempt to phenocopy met1 mutants. In Arabidopsis, three Fluorescent Tagged Lines (FTL) were used to visualise recombination between two markers directly in the seeds. These markers spanned peri-centromeric, interstitial and sub-telomeric loci on the chromosomes. Changes in recombination were observed in the sub-telomeric interval but not in the interstitial centromeric intervals. Additionally, gene expression was measured using RT-qPCR to assess the general effect of zebularine on plant development, showing significant changes in gene regulation in the presence of zebularine in genes involved in germination, vegetative/flowering stage balance, stress response and DNA-methylation. In barley, zebularine treatment triggered delayed development during germination but the plants quickly recovered, and no effect was observed on the recombination landscape when measured using SNP genotyping. Gene expression was also analysed using microarray, but the effect of the zebularine treatment was too stochastic on the seedlings to generate conclusions on its effect on gene regulation. In parallel, a large TILLING (Targeting Induced Local Lesions in Genomes) population was used to identify a collection of barley met1 mutants (cv. Golden Promise). Two lines carrying missense mutations were identified where the nucleotide change is predicted to have a highly deleterious effect on protein function. These lines were crossed into another WT background (cv. Barke) and F3 families were generated. The plants were characterized for effects on plant performance and were then genotyped using a 50k SNP genotyping array. F3 genotyping shows a small tendency to redistributed recombination frequency in met1 mutants in a non-significant manner. Finally, a CRISPR-Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats) construct was developed to target mutagenesis in met1 and ddm1 in barley cv. Golden Promise.