Horizontal gene transfer of viral inserts from GM plants to viruses
It is our conclusion that too little is known at present about the evolution, ecology, biochemistry and pathogenicity of viruses to allay the concerns about horizontal gene transfer (HGT) from plants containing viral transgenes. Risk assessment of transgenic crops for HGT involves extrapolation to large scales and diverse real-world situations, which cannot at present be done without the use of crucial and unsubstantiated assumptions. For example, it is a fundamental tenet of risk assessment that extrapolation from limited knowledge requires a robust mechanistic understanding of a new technology. Failure to establish this basic information can lead to incorporation of false assumptions into risk-assessment procedures. In the case of virus-resistant transgenic crops this information should certainly include substantial knowledge of the mechanism of virus-resistance, its characteristics and limitations. For instance, it now seems likely that the resistance mechanism for virus-resistant lines is via post transcriptional gene silencing (PTGS) and it is now known that PTGS is turned off by many viruses (e.g. Voinnet et al 1999). Thus a key assumption relied on by the USDA to argue against the likelihood of HGT now appears to be invalid (USDA 1994). As we showed in this paper, this was not the only unvalidated assumption used to justify approval. Not only is it unscientific to rely on unvalidated assumptions it is also indefensible to at the same time argue (before the general public), that only well-understood transgenes and systems are used in the manufacture of transgenic crop varieties.
We are not alone in concluding that more information on plant virus biology is needed and many authors have argued the necessity for more information (de Zoeten 1991; Tepfer 1993; AIBS 1995; Miller et al 1997; Aaziz and Tepfer 1999; Power 2002; Tepfer 2002). None of these authors has however made explicit the link between inadequate baseline information and irresponsible regulation. Unlike the USDA and other national regulatory bodies, we believe that lack of such knowledge requires delaying applications and withdrawing approval for the release of GM virus sequence-containing plant varieties on the grounds that approval is incompatible with a prudent and precautionary approach.
Furthermore, most of the HGT risks carried by commercial transgenic varieties containing virus inserts are probably unnecessary ones. To reduce HGT, non-viral promoters could be used, transgene sizes reduced, replication origins removed, protein expression prevented and gene sequences disabled. These improvements would probably not impact on transgene effectiveness. By not requiring applicants to incorporate these safety features, or to justify why they have not done so, regulators are exposing third parties to unnecessary risks.
The question of monitoring the consequences of virus sequences in GM crops is crucial. The power of science derives from testing and refining its predictions. If we do not test predictions of a lack of impact of viral sequences in GM crops we will either learn nothing about the risks and hazards or we will learn it too late.
What is required above all is for the scientific community to make itself heard so that poor biosafety assessments are challenged in print and best practice is incorporated into commercial varieties. By not doing this the scientific community is perceived as failing to be independent of commercial pressures. It is difficult to see however, individual members of the scientific community, who wished to express such views, being able to do so explicitly and openly in the present climate.