Parental genomic imprinting (PGI) is the selective expression of only one allele after fertilization depending solely on its parent-of-origin. In Arabidopsis, PGI has been characterized in the seed endosperm linking this phenomenon to the provision of maternal resources. Understanding the developmental programs regulated by PGI may provide clues to optimizing breeding strategies for seed size.
Loss of PGI can result in large seeds when maternal silencing is compromised or small seeds when paternal silencing is compromised. In a typical breeding scheme, it might be expected that PGI could lead to parent-of-origin effects or specific maternal/paternal interactions that would be masked as a component of heterosis. For example, the Landsberg erecta (Ler) ascension of Arabidopsis has a small seed size, but Ler pollen sires large seeds on a Cape Verde Islands (Cvi) mother. Dissecting these epigenetic effects from heterosis or simple additive effects requires a robust system to examine genome-wide genetic contributions to specific phenotypic effects. This is distinct from the more limited analysis of mechanisms for imprinted gene expression that have been characterized extensively in select genes.
To ascertain the adaptive role of global imprinted gene programs in wild-type plants we are simultaneously assaying the genetic basis for parent of origin effects Ler, Cvi and other ascensions, while exploring the cellular basis for differences in early development. A parental bias is observed in these lines with Ler fathers promoting a transgressive large seed not seen in reciprocal crosses. Using spectral analysis of autofluorescence in the seed, we isolated endosperm from surrounding maternal structures in non-transgenic natural variants. Reconstructed 3D endosperm models show that Cvi development involves an adaxial-abaxial expansion not present in Ler, which restricts expansion on this axis maternally. Ler fathers, however, promote an anterior-posterior expansion.
Using recombinant inbred lines, we are further tracking the parental origins and loci associated with changes in early seed development. Previously characterized QTL have been linked to paternal or maternal effects. Further, geometric morphometrics have been applied to seed shape to better discern if there are independent genetic pathways responsible for aspect of final seed size. We are currently linking these final seed size outcomes to events in early development to better understand the developmental targets of seed size adaptations and the role of PGI.