Four GM plant species (Arabidopsis thaliana, Brassica napus, Nicotiana benthamiana and N. tabacum), each expressing the gene encoding the jellyfish green fluorescent protein (GFP) regulated by the cauliflower mosaic virus (CaMV) 35S RNA promoter, were assessed for the extent of transgene silencing and viral genome integration following infection by CaMV. The first three species are systemic hosts of CaMV, while N. tabacum is only a local host for a few strains of CaMV. A generalized systemic silencing of the GFP transgene was not observed in a total of 100 plants of each species infected with CaMV, although some localized loss of GFP was observed in CaMV-infected N. benthamiana leaves, and some loss of fluorescence was observed in older leaves of uninfected as well as infected plants. Progeny seedlings obtained from the above infected plants also did not exhibit transgene silencing showing that virus infection did not affect the stability of the transgene. These progeny plants also did not show signs of virus infection, indicating that the presence of the CaMV 35S RNA promoter sequences in the plant genome did not potentiate seed transmission of the virus. Integration of infective CaMV into the CaMV 35S RNA promoter could not be detected in 944 samples taken from leaves of the above infected plant species or in 2912 samples taken from progeny seedlings. Based on a detection limit of one copy per 4000 haploid genomes, we conclude that if integration of virus does occur into the CaMV 35S RNA promoter, then it occurs at such a low frequency as to be insignificant.

We report a survey of four viruses (beet western yellows luteovirus (BWYV), cauliflower mosaic caulimovirus (CaMV), turnip mosaic potyvirus (TuMV), turnip yellow mosaic tymovirus (TYMV)) in five natural populations of Brassica oleracea in Dorset (UK). All four viruses were common; 43% of plants were infected with BWYV, 60% with CaMV, 43% with TuMV and 18% with TYMV. For each virus there were significant differences in the proportion of infected plants among populations, which were not completely explained by differences in the age of plants. Multiple virus infections were prevalent, with 54% of plants having two or more virus types. There were statistically significant associations between pairs of viruses. The CaMV was positively associated with the other three viruses, and BWYV was also positively associated with TuMV. There was no detectable association between BWYV and TYMV, whereas TuMV and TYMV were negatively associated. We suggest these associations result from BWYV, CaMV and TuMV having aphid vectors in common, as aphids are attracted to plants that already have a virus infection. Infected plants were distributed randomly or were very weakly aggregated within populations. The implications of widespread multiple virus infections in natural plant populations are discussed with respect to the release of transgenic plants expressing virus-derived genes.