Gene expression landscape of the Brassica napus seed reveals subgenome bias in both space and time

Brassica napus (canola; AnAnCnCn) contains both complete diploid genomes from its progenitors B. rapa (An) and B. oleracea (Cn). Despite growing knowledge of the gene expression landscape of the B. napus seed, little is known about subgenome bias underpinning the development of specific cells and tissues across the seed lifecycle. Here, we present a large-scale transcriptome atlas of the B. napus seed, including both the maternal seed coat and filial embryo and endosperm subregions. We report on extensive, global Cn subgenome bias throughout development and use homoeologous gene pairs to describe how subgenomic bias differs across subregions. We find that subgenome bias is most prominent during early development and that the maternal subregions experience far more asymmetric transcript accumulation in favour of the Cn subgenome. In particular, the unexpectedly distinct transcriptome profile of the chalazal pole indicates the unique developmental processes involved within the chalaza. Further, we report that genes integral to seed storage comprise a large portion of the transcriptome of mature seeds, especially within the embryo, and that gene pairs previously documented to be instrumental in seed development exhibit low transcriptional bias. This work represents an important synthesis of polyploid transcriptomics in seed biology and provides a comprehensive overview of the B. napus gene expression landscape in both space and time.