*For offprint requests

Key words: RNA recombination, plant viruses, bromoviruses

Each year plant viral diseases significantly reduce world wide food production. The traditional long term approach of protecting economically important plants from viral diseases has been to identify viral resistance genes within wild relatives and move those genes into the commercial cultivars by traditional breeding techniques. Unfortunately the supply of such genes is limited and it has become increasingly difficult to find new naturally occurring resistance genes among the wild relatives of important crop plants. Thus alternative methods of providing virus resistance have become attractive.

The genetic modification of plant genomes by the techniques of recombinant DNA provides a method for the introduction of genes into plants from unrelated sources. The transfer of known viral resistance genes from one plant species to another awaits further characterization of such genes and testing of the resistance mechanism in different genetic backgrounds. Curiously, a current source of useful resistance genes is found within the genomes of the viruses themselves. Expression of one of several different plant virus genes within a plant provides resistance to the virus from which the gene was derived. The commercial development of such virus resistant transgenic plants is well underway.

As with all new technologies, evaluation of the risks involved with the commercialization and environmental release of a new product is wise. Although it may be difficult to foresee all possible risks, sincere attempts should be made to eliminate or reduce potential hazards and ensure that the benefits of the product outweigh the risks. Evaluation of virus resistant transgenic plants is no exception.

With this in mind, we have focused our attention on the possibility that the segment of the viral genome expressed in the virus resistant transgenic plant may be incorporated into the genome of other viruses that may challenge the transgenic plant in the field. RNA recombination is a documented evolutionary mechanism of RNA viruses; it appears to depend on the error prone viral RNA polymerase that occasionally changes templates during the replication of an RNA virus. The result of such a switch is that RNA from a distinctly different template is incorporated into a copy of the viral genome. Such an event provides a chimeric viral RNA that is subjected to natural selection. The role of recombination in the evolution of RNA viruses has been suggested by the nucleotide sequences of several RNA viruses (Edwards et al., 1992; van der Kuyl et al., 1991). Additionally, RNA recombination has been demonstrated and studied in numerous laboratory experiments.

Our recent experiments have dealt with the possibility that the viral transgene of the transgenic plant can be incorporated into a challenging virus. Using cowpea chlorotic mottle bromovirus (CCMV) as a model system, we recovered viable recombinant virus from 3% of the transgenic Nicotiana benthamiana plants that we inoculated (Greene and Allison, 1994). In these experiments transgenic plants were constructed that expressed the 3' two thirds of the capsid gene plus the associated 3' untranslated region (UTR). Transcription of the viral transgene was inadequate to provide CCMV resistance. These plants were inoculated with a CCMV deletion mutant which lacked the 3' third of the capsid gene. While disruption of the capsid gene inhibited the systemic movement of the virus, it did not prevent virus replication. Consequently, any recombination events that successfully restored a functional CCMV capsid gene resulted in the systemic infection of the transgenic plant. Evaluation of the systemically infecting virus confirmed the incorporation of the transgene within the recombinant virus. These experiments and others which have used a similar approach but different plant/viral systems (Lommel and Xiong, 1991; Gal et al., 1992; Schoelz and Wintermantel, 1993) have demonstrated the availability of the viral transgene to the replicating virus.

In our experiments the CCMV transgene included the 3' UTR that is normally associated with the capsid gene in the wild type virus. This segment of RNA is highly structured and folds into a tRNA-like structure that is aminoacylated by host enzymes (Ahlquist et al., 1981). This region also includes recognition sequences for the initiation of RNA replication (French and Ahlquist, 1987; Dreher and Hall, 1988) which begins at the 3' terminus of the RNA. The following experiments suggested that the presence of the complete 3' UTR within the transgene enhanced the involvement of the transgene in recombination events (Greene and Allison, 1996). Transgenic plants were constructed in which the CCMV transgene was similar to that described above but lacked 69, 83 or 214 nucleotides of the 3' terminal sequence. These plants, which were not resistant to the wild type virus, were challenged with the same inoculum that had yielded viable recombinants in the previous experiment (Greene and Allison, 1994 ). No systemic infections resulted in the 479 plants that were inoculated. This was in contrast to the recovery of seven recombinant viruses from the inoculation of 235 transgenic plants which transcribed the same coding region plus the complete 3' UTR . Consequently, in this bromovirus system these terminal deletions reduced the recovery of viable recombinants.

There are several possible reasons why the larger transcript may be more likely to be involved in RNA recombination events. One, the larger transgenic transcript may simply provide a larger target for RNA recombination. Two, the complete 3' UTR may mimic the wild type CCMV 3' terminus by folding to form the 3' natural CCMV tRNA-like structure. This terminal structure may provide the transcript some protection from RNAses and, in comparison to the deleted versions, extend the cytoplasmic half-life of the transcript. Such an extended life may make the transcript with the complete 3' UTR more available for recombination than the deleted version. Additionally, a transcript with a complete 3' UTR may mimic the wild type RNA and be recognized by the viral replication complex. Thus replication could initiate on the transgenic transcript. If the replication complex completes replication of the transgene, a complementary copy, minus strand, of the transgene would be produced and it too would be available for recombination during plus strand synthesis. Preliminary results indicate that a minus strand copy of the CCMV transgene is produced when the complete 3' UTR is present. Therefore, avoiding the 3' UTR in transgenes or disrupting this region in an attempt to mutate the replicase binding site may reduce the frequency in which the transgene is involved in recombination. While these measures may reduce the recombination potential, they do not eliminate it.

Although these results and those from other labs indicate that the transgene is available for recombination with a replicating virus, transgenic plants do not necessarily provide recombination opportunities that are entirely new to plant viruses. There are numerous reports of mixed infections where two or more viruses are observed in the same plant at the same time. This situation would appear to provide for unlimited recombination opportunities. But the presence of two viruses in the same plant at the time of sampling does not necessarily mean that they have undergone replication simultaneously within the same cells, in fact many mixed infections may reflect an accumulation of viruses over an extended period of time. During the infection process, viral replication advances through the plant tissues initiating and likely completing replication as it progresses (Wang and Maule, 1995). Thus two viruses of a mixed infection that were not inoculated simultaneously may have limited recombination opportunities. But surely there are instances where recombination could occur, particularly in the cases of viral complexes in which the infection of one virus is dependent on the presence of another. Our current research with mixed infections of bromoviruses indicates simultaneous mixed infections are difficult to establish. Once established, however, both viable and non-viable recombinant products are recovered from mixed infections that involve bromovirus capsid protein mutants.

In contrast to mixed infections, transgene transcription is controlled by constitutive promoters, such as the 35S promoter, and transgenic transcripts are available continually within the cytoplasm of every plant cell. Thus replicating RNAs of a challenging virus will have access to transgenic transcripts. Consequently, constitutive expression of the transgene in virus resistant transgenic plants appears to provide unique opportunities for recombination that are not mimicked in natural mixed infections.

REFERENCES

Ahlquist, P. Dasgupta R. Kaesberg P. 1981. Near identity of 3' RNA secondary structure in bromoviruses and cucumber mosaic virus. Cell 23: 183-198.

Dreher TW, Hall TC. 1988. Mutational analysis of the sequence and structural requirements in brome mosaic virus RNA for minus strand promoter activity. J. Mol. Bio. 201: 31-40.

Edwards, RC, Petty ITD, Jackson AO. 1992. RNA recombination in the genome of barley stripe mosaic virus. Virology 189: 389-392.

French R, Ahlquist P. 1987. Intercistronic as well as terminal sequences are required for efficient amplification of brome mosaic virus RNA 3. J. Virol. 61: 1457-1465.

Gla S, Pisan B, Hohn T, Grimsley N, Hohn B. 1992. Agroinfection of transgenic plants leads to viable cauliflower mosaic virus by intermolecular recombination. Virology 187: 525-533.

Greene AE, Allison RF. 1994. Recombination between viral RNA and transgenic plant transcripts. Science 263: 1423-1425.

Greene AE, Allison RF. 1996. Deletions in the 3' untranslated region of cowpea chlorotic mottle virus transgene reduce recovery of recombinant viruses in transgenic plants. Virology 225: 231-234.

Lommel SA, Xiong Z. 1991. Reconstitution of a functional red clover necrotic mosaic virus by recombinational rescue of the cell-to-cell movement gene expressed in a transgenic plant. J. Cell Biochem. 15A:151.

Schoelz JE, Wintermantel WM. 1993. Expansion of viral host range through complementation and recombination in transgenic plants. Plant Cell 5: 1669-1679.

van der Kuyl AC, Neeleman L, Bol JF. 1991. Complementation and recombination between alfalfa mosaic virus RNA 3 mutants in tobacco plants Virology 183: 731-738.

Wang D, Maule AJ. 1995. Inhibition of host gene expression associated with plant virus replication. Science 267: 229-231.