Analysis of the effects of high selection pressure on recombination between RNA3 of Cucumber mosaic virus and transgene mRNAs

It has been shown that mixed infections (with two or more viruses) occur frequently in nature, and that when two related viruses replicate simultaneously in the same cells, genetic recombination between them can take place. There is also evidence that recombination can occur between cellular and viral RNAs. Thus, virus-resistant transgenic plants (VRTPs) that express viral sequences could be a source of novel recombinant viral genomes. From a biosafety point of view, it is therefore important to understand if the recombinants that are generated in VRTPs are novel, and if they could contribute to an increased risk of emergence of recombinant viruses. Since it is impossible to test all possible recombinants, the best way to evaluate this risk is to compare the recombinants that appear in infected transgenic plants with those that appear in doubly infected non-transgenic plants under conditions of low selection pressure. Cucumber mosaic virus (CMV) is the most important member of the genus Cucumovirus, family Bromoviridae. As with all Bromoviridae, the CMV genome consists of three plus-sense, single-stranded RNA molecules (RNA1, RNA2, and RNA3). CMV isolates are divided in two main subgroups (subgroup I and subgroup II), distinguished by serological and nucleic acid properties. Infecting over 800 plant species, CMV is considered one of the most important plant viruses, causing serious economically diseases in many parts of the world. Using a low selection pressure approach, colleagues in the lab had previously shown that in tobacco plants infected with two strains of CMV, a subgroup I strain (I17F-CMV) and subgroup II strain (R-CMV), the populations of recombinant viral RNAs found were equivalent to those found in tobacco transgenic for the entire coat protein (CP) open reading frame and the 3’ non-coding region (3’NCR) of R-CMV infected with I17F-CMV (Turturo et al. 2008). These results suggest that in this case novel viral recombinants are not expected to appear. But, since the detection of host-messenger/viral RNA recombinants under conditions of low selection pressure is difficult because of the low number of recombinant molecules in relation to parental molecules, it was decided to analyze the recombinants appearing under conditions of high selection pressure in favor of recombinants. To do this, we inoculated 44 plants of the same line expressing the R-CMV CP+3'NCR transgene with in vitro transcripts of a mutant I17F-CMV that had been attenuated by a 6-nt deletion in the 3'NCR of RNA3. Samples were taken from symptomatic plants 8, 10, and 15 days post-inoculation. As for the studies under low selection pressure, plants were screened for recombinant RNA3 by RT-PCR, and the recombinant nature of the amplicons was verified by cloning and sequencing. Many of the recombinants were different from those observed in the same plant line when inoculated with wild-type I17F-CMV in the absence of selection pressure. Further studies will determine the abundance of the different recombinants, and will aim to clarify why they were not previously detected.