The poor lonesome Brassica napus A subgenome may not survive without its mate

International audienceWhole genome duplication (WGD) is recognized as a major force in plant evolution and speciation, resulting from both structural and/or functional alterations. These modifications occur in the first few generations following WGD events and continue during the lifespan of the polyploid species. To study the long-term evolutionary impact of allopolyploidy, previous studies have almost exclusively compared a present allotetraploid species with its current diploid progenitors. However, the diploid progenitors have independently evolved since the allotetraploid formation, preventing to fully understand its evolution. To circumvent this problem, another approach corresponding to the subgenome extraction, offers the unique opportunity to elucidate the long-term evolution in an allopolyploid context. Presently, one of the best model to determine the role of allopolyploidy in genome evolution corresponds to oilseed rape (Brassica napus, AACC, 2n=38), formed about 8,500 years ago from a cross between B. rapa (AA, 2n=20) and B. oleracea (CC, 2n=18). To identify the structural rearrangements that occurred since this allotetraploid formation, we performed two different strategies to extract its diploid AA component. For the 1st strategy, a cross between B. napus var. Darmor and B. rapa was realized to produce a triploid AAC F1 interspecific hybrid. This triploid was then backcrossed three times to B. napus and the AAC progenies were selected at each generation in order to obtain an almost pure B. napus A subgenome. We finally attempted to separate the AA and C components by selfing the last AAC plant but no AA plant was obtained. For the 2nd strategy, a cross between the initial AAC F1 hybrid and B. rapa was performed to generate AA plants. After four cycles of selfed and also backcrossed to B. napus, AA plants mainly containing the B. napus A subgenome were obtained. Using the 60k SNP Illumina micro array and the genome sequence of B. napus var. Darmor, we assessed the genomic structure of the last AA plants produced. They presented a lower proportion of B. napus A subgenome extracted than expected and some introgressions from the C subgenome. Our analyses revealed that the genomic regions from B. rapa conserved in diploid AA plants were not randomly distributed along the A subgenome, suggesting that the A subgenome may not survive without the C subgenome, most presumably due to the loss of one homoeologous copy since the formation of the allotetraploid B. napus