Introgression of crop genes from oilseed rape (Brassica napus) to related wild species - An avenue for the escape of engineered genes

The risk of transgene transfer from genetically modified oilseed rape to the weedy relatives Brassica rapa and B. juncea has been studied. In field experiments with mixed populations of B. napus and B. juncea the number of hybrid seeds produced spontaneously by B. juncea was dependent on the proportion between the two species; the hybrid production with B. napus as the female was much less responsive to changes in the species proportion. Hybridisation between B. napus and B. rapa was found to be dependent on the spatial structure of crop and recipient species as well as the genotype of the parents in the cross. A combined estimate of fitness showed that the Fl hybrids between oilseed rape and B. rapa were more fit than B. rapa. The Fl hybrids backcrossed spontaneously to the weedy parent in field experiments. Some of the BC1 plants from this cross displayed a B. rapa-like morphology, and they had a chromosome number of 2n=20 (as B. rapa) and a high fertility. Transgenes were transferred from the genetically engineered B. napus parent and expressed in the backcross plants. In general fitness of BC1 and Fz plants from crosses between oilseed rape and B. rapa had reduced fitness compared to the pure species, however, some of these introgressed plants were as fit as B. rapa or B. napus. An analysis of genetic variation in populations of wild and cultivated B. rapa, cultivars of B. oleracea and B. napus indicated that introgression of oilseed rape genetic material to B. rapa occurred under certain agricultural conditions. Unlike the weedy B. rapa parent none of the hybrids B. napus x B rapa and B. rapa x B. napus revealed seed dormancy. A study of the transmission of B. napus specific DNA-markers to the BC1 generations (with B. rapa or B. juncea as recurrent parent) showed that all analysed markers were transferred. The vast majority of these B, napus markers were located to chromosomes of the C-genome indicating that a targeted integration of transgenes to this genome will not exclude transfer of the genes in interspecific crosses. In conclusion, our results show that there is a risk that transgenes will be transferred from oilseed rape to B. rapa and B. juncea, in particular at locations that favour survival of the weedy populations and their interspecific hybrids with oilseed rape (e.g. fallow land).