Genomic imprinting leads to unequal expression of genetic alleles in mammals and angiosperms and is present only in the endosperm in angiosperms. The endosperm of angiosperms is formed by the fusion of male and female cells and differentiates into terminal tissues. The timing of cellularization in the development of the endosperm is altered in interploidy or interspecific sexual hybridization, affecting endosperm development and resulting in seed inviability, and imprinted genes play a role in maintaining hybrid barriers.
Genomic imprinting involves histone modifications and DNA methylation. Genome-wide epigenetic remodeling establishes imprinting at hundreds of loci and may play a role in maternal suppression and paternal promotion of endosperm development through imprinted genes. Genome-wide imprinted gene expression has been found in many plant species, including rice. However, current understanding of epigenetic control and temporal patterns of gene expression is limited to well-characterized imprinted genes.
Recently, Nature Plants published a research paper entitled "Multilayered epigenetic control of persistent and stage-specific imprinted genes in rice endosperm". This study classified persistent and stage-specific imprinted genes based on transcriptome analysis of different developmental stages of rice endosperm. The overall trend showed that the maternal alleles of maternally expressed imprinted genes were activated by DNA demethylation, while the maternal alleles of paternally expressed gene body methylation (gbM) imprinted genes were silenced by DNA demethylation and H3K27me3 deposition. These regions are associated with an enriched motif associated with Tc/Mar-Stowaway. This finding provides insights into the stability of genomic imprinting and potential variations associated with endosperm development, different cell types, and parental genotypes.
This study performed RNA-seq analysis at different stages of endosperm development, using endosperm 7 days after pollination as a control, and repeated two replicates using endosperm from two reverse hybrid combinations. The results showed that the trend of unequal expression of imprinted genes was conserved in the tested genetic combinations, but there was no overlap in the two crosses, i.e., there were few imprinted genes except for genes expressed early in embryonic development.
This study identified more than 100 stage-specific imprinted genes and many previously uncharacterized genes that were imprinted during rice endosperm development. And to understand the different categories of imprinted genes associated with specific biological functions, gene ontology (GO) analysis was performed. Some of these genes are essential for endosperm development but have not been identified as imprinted genes before. Although the roles of many of the newly discovered persistent and stage-specific imprinted genes have not yet been evaluated, these genes may regulate gene expression during endosperm development.
To investigate the epigenetic control of persistent and stage-specific MEGs and PEGs, allele-specific whole-genome bisulfite sequencing was performed on F1 embryos and endosperm. CG methylation was found to be different between maternally and paternally inherited alleles, and persistent MEGs had higher methylation levels in paternally inherited alleles. However, MEGs are not an absolute imprinted gene class, and their expression is correlated with paternal DNA methylation levels. CG and CHG DNA methylation differences were not associated with PEG paternal bias, suggesting that other epigenetic mechanisms may affect PEG imprinting.
The study further analyzed differentially methylated regions (DMRs) and focused on the H3K27me3 mark, which was found to be associated with paternally biased expression. Differences in H3K27me3 allelic accumulation between persistent and stage-specific imprinted genes require further evaluation. Studying cell-specific expression patterns of imprinted genes by snRNA-seq revealed cell type-specific and parental unequal expression of persistent and stage-specific MEGs and PEGs.
This study used the MEME program to search for cis-regulatory motifs associated with DNA demethylation or H3K27me3 enrichment and found specific motifs associated with h3k27me3-enriched regions. This motif is significantly enriched in Tc/Mariner-Stowaway mini-inverted repeats, suggesting that PRC2 targets demethylated DNA regions. H3K27me3 differs between peg and other genes, but the demethylated Stowaway motif is not the only relevant cis motif. Studying the relationship of imprinted genes with epigenetic marks and stowaway motifs suggests that DNA demethylation and H3K27me3 deposition may establish PEG by repressing maternal allele expression.
This study's genome-wide time-course analysis of rice allele expression revealed two classes of imprinted genes - persistent and stage-specific - and multiple layers of epigenetic control of endosperm genomic imprinting. MEGs showed demethylation of CG/CHG sequences, resulting in endosperm-specific expression. In contrast, PEGs often showed gbM and were expressed in multiple tissues. Most gbM PEGs underwent maternal allele-specific demethylation. DNA demethylation may be a prerequisite for the establishment of PEG genomic imprinting, and H3K27me3 deposition may occur subsequently. Silent persistent imprinted genes had DNA methylation and H3K27me3 marks on both MEGs and PEGs, while stage-specific imprinted genes showed more abundant modifications and their expression varied during endosperm development. Although the epigenetic states of these genes are not discontinuous, their existence raises the question of their role in endosperm development and evolution, and further identification may provide new ideas for the evolution and regulation of rice endosperm development.
