Abstract
Medea, the often-misunderstood Greek mythological figure infamous for killing her own offspring, influenced the naming of the first imprinted gene identified in the model organism Arabidopsis thaliana. MEDEA (MEA) shows exclusive maternal expression and is required for normal seed development. Maternal inheritance of a mea mutant allele leads to seed abortion, while there is no effect on seed viability when the mea allele is transmitted paternally. MEA is regulated by genomic imprinting, a form of epigenetic gene regulation that results in allele-specific expression depending on the parental origin. This phenomenon occurs in both animals and plants, but the molecular mechanisms of imprinting are not as well characterised in the latter. Little is known about plant imprinting control regions (ICRs), the cis-elements that are required and sufficient to mediate imprinted expression. At the MEA locus, an ICR was localised to the upstream 200 base-pair (bp) cis-regulatory region (MEA-ICR). However, the activity of this ICR is independent of DNA methylation and histone H3 Lysine 27 trimethylation (H3K27me3), both epigenetic marks typically associated with imprinting. Thus, this study aims to unravel the molecular mechanism underlying the imprinting regulation of MEA by characterising the MEA-ICR. Using multi-gRNA CRISPR/Cas9 mutagenesis, we generated and analysed a series of mutant MEA-ICR alleles for their effects on seed development. Mutations in certain regions within the MEA-ICR led to maternal-effect seed abortions, with a 15 bp region found to be particularly detrimental. Further characterisation of this 15 bp region using a 3xVenus-NLS reporter system determined that it is required for MEA imprinted expression. Deletion of this 15 bp region from the MEA-ICR led to loss of imprinting but not expression of the transgene, allowing MEA imprinting to be uncoupled from transcription for the first time. Additionally, through yeast-1-hybrid assays, we confirmed the interactions of several putative trans-acting factors with the MEA-ICR, including the central cell-specific histone H3 variant HTR14. Therefore, we propose a new working model for the regulation of MEA imprinting that likely involves a nucleosome at the MEA-ICR. In summary, this study has unveiled a novel, methylation-independent imprinting control element in plants, bringing us one step closer to elucidating the primary imprint at the MEA locus. Our findings may provide insights into other noncanonical imprinting mechanism in both plants and animals. This is important as the dysregulation of some imprinted genes is associated with human diseases.