Should not be approved, major gaps in risk assessment Prof. Joe Cummins and Dr. Mae-Wan Ho
This report was submitted to the USDA on behalf of the Independent Science Panel
The United Sates Department of Agriculture (USDA) has announced that its Animal and Plant Health Inspection Service (APHIS) has received a petition from its Agricultural Research Service (ARS) seeking non-regulated status for a transgenic plum designated transformation event C5, genetically engineered to resist infection by plum poxvirus (PPV), and is soliciting public comments on whether this plum presents a plant pest risk. The closing date for making public comment is 17 July 2006 (http://www.regulations.gov/fdmspublic/component/main)
It is worth mentioning that the transgenic plum petition is the first temperate transgenic tree to be petitioned for non-regulated status. Petitions for a number of transgenic trees are certain to follow in short order including transgenic forest trees, which would be really disastrous for the world’s forests [1] (GM forest trees, the ultimate threat, SiS 26). Those petitions will be of low quality unless sufficient public participation is encountered in this first petition.
So please enter your objections now. Use this article freely.
A version of the same petition was first submitted to the USDA in 2004, and the current petition open for public comment is a revised version submitted in March 2006 [2], together with an updated environment assessment [3]. The most salient feature of the revised petition and assessment is that the gene for the viral coat protein was found not to produce a viral protein but to initiate a process called post-transcriptional gene silencing associated with a small inhibitory RNA, a short sequence of RNA which can be used to silence gene expression.
The proposed commercial release is the patented plum variety “Honey Sweet” plum developed jointly by USDA, Institut National de la Recherche Agronomique, Paris, France and Cornell University. The plum tree has the plum poxvirus (PPV) coat protein gene incorporated to provide resistance to the major plum pest PPV. The female parent of the plum is “Bluebyrd” (named for Senator Robert Byrd), while the pollen parent is “unknown”. The plant is not self fertile, a pollinator is required. The variety is propagated by bud grafting to standard rootstocks [3]. The plum fruit is a typical drupe in which the skin and flesh of the fruit contain only maternal genes; the seed embryo and endosperm contain both paternal and maternal genes. The seeds of the transgenic plum are viable and could produce viable plants. A non-transgenic plum tree pollinated by the transgenic plum will give fruits that will not contain the PPV gene in their flesh, only the seed would. All the plums produced on transgenic tree, however, fruit and seed, would be transgenic, regardless of the status of the pollinator.
The transgenic plum contains the PPV coat protein gene along with the selectable markers NPTII (Kanamycin resistance) and GUS (β-Glucuronidase). There are multiple copies of the PPV coat protein gene linked at the insertion site. The genetic modification of the plums was done using a gene cassette containing the NPTII gene driven by the relatively weak nos promoter from Agrobacterium and terminated by the nos terminator. The PPV-CP was driven by the cauliflower mosaic virus (CaMV) promoter and transcription was terminated by the nos terminator from Agrobacterium. The GUS gene was also driven by the CaMV promoter and transcription terminated with nos. Analysis of the genes inserted into the plum clone C5 showed that there was a second PPV gene insert unlinked to the primary NPII, GUS, PPV-CV gene insert. Fragments of the NPII gene and the GUS gene were also detected in the transgenic plum. The multi-copy PPV inserts appeared to behave like a single gene in crosses, indicating that they are relatively close together on a chromosome.
There were no attempts to characterize the inserts with fluorescent in situ hybridization of chromosomes and to identify genome sequences flanking the insert(s), both of which are now standard.
Sequencing was done, and results were “inconclusive” on account of “transgene duplications, rearrangements, and …an inverted repeat of the PPV-CP gene...” [2, p. 29] Sequences from the plasmid vector pBR322 were also present in the insert(s).
PPV-CV insert was apparently methylated and its CaMV promoter more lightly so, unlike the promoter or the GUS gene [2]. Nevertheless, the PPV-CV is actively transcribed, and then presumably degraded in post-transcriptional gene silencing, so only low levels of mRNA accumulated.
Post-transcriptional gene silencing is a sequence-specific post-transcriptional RNA degrading system that is programmed by the transgene-encoded RNA sequence [5].
The insertion of the PPV-CP gene cassette into the plum is necessary but not sufficient to produce strong stable resistance to PPV. For example plum transformation events C2, C3 and C4 accumulated high levels of PPV-CP messenger RNA and coat protein but were not resistant to PPV. In contrast, event C5 produced little PPV-CP messenger RNA and barely detectable coat protein [6], but its PPV resistance appeared stable in open field trials together with controls either without transgenes or with other transformation events that were not virus resistant such as event C3 [2, 7]. The stable viral resistance of C5 was associated with the duplication and methylation of the PPV-CP gene, and also with a small RNA species present in high concentrations (see later).
One potential problem with the C5 event released into the environment is the transfer of the NPTII kanamycin resistance gene to soil bacteria and in turn, to animal pathogens. The NPTII gene was extensively transferred from transgenic sugar beet to a soil bacterium, Actinobacter, in an experimental situation [8]. Even though the rootstock for the C5 plum is not transgenic and not able to transfer the NPII gene, the autumn leaves, shed bark and flowers of the plum would certainly deliver a good quantity of the antibiotic resistance gene to the soil. Sequence homologies to bacterial sequences, including GUS gene and nos terminator and pBR322 plasmid, are expected to greatly increase the frequency of horizontal gene transfer, up to a billion-fold [9]. Furthermore, the horizontal transfer of non-homologous DNA occurs at relatively high frequencies when a homologous DNA ‘anchor sequence’ is present, which can be as short as 99bp. There are at least 87 species of naturally transformable bacteria in the soil [10]. As trees are long-lived, there is every opportunity for horizontal gene transfer to take place from transgenic trees.
In 2004, the United States EPA published a Scientific Advisory Panel (SAP) Report on Plant Incorporated Protectant, specifically those based on viral coat proteins (PVCP-PIPS) [11]. The report provided extensive discussion of concerns such as the spread of virus resistance to weedy relatives, but did not deal with the implications of post-transcriptional gene silencing, horizontal gene transfer, or viral interactions in the wild.
We have drawn attention to the recombination between viral transgenes and invading viruses in connection with the hazards of the cauliflower mosaic virus 35S promoter [12] (Hazards of transgenic plants containing the cauliflower mosaic viral promoter) that is in practically every transgenic plant commercially grown, and is present in the transgenic plum, driving both the PPV-CP gene and the GUS gene. We pointed out that as the CaMV 35S promoter contains a recombination hotspot, it is more likely to take part in horizontal gene transfer recombination and is a major cause of transgene instability. This prediction has been confirmed since [13-15] (Transgenic lines proven unstable; SiS 20, Unstable transgenic lines illegal, SiS 21). Five out of five transgenic lines commercially approved had rearrangements of the transgenic insert, and the CaMV 35S promoter was a frequent breakpoint. The issue of transgenic instability remains unresolved to this day, and we are not convinced that the transgenic plum petitioned for non-regulated status is stable in the absence of the appropriate molecular genetic data. In fact, there are signs that it too is unstable (see later).
There are other ways in which viruses interact: heterologous encapsidation (the transgenic coat protein adding to the capsid of an unrelated invading virus and therefore helping it escape inactivation by the host, and synergy, in which invading viruses supply suppressors of post-transcriptional gene silencing mounted by the host, therefore cancelling out the viral resistance. The SAP believed that heterologous encapsidation and synergy were relatively unimportant in PVPCP-PIPS [11] and felt that recombination could be prevented by removing the untranslated tail end of the gene construct, even though there was limited support for that supposition. The panel concluded that eating transgenic viral coat protein should be considered safe (without experimental verification) because people have been eating virus infected plant material for a long time. We have argued in detail why that assumption is invalid in the case of viral DNA such as the coat protein gene or the CaMV 35S promoter [12], basically because a viral gene isolated and placed in a foreign genetic and evolutionary context can never be equated with the gene in the natural virus (Hazards of transgenic plants containing the cauliflower mosaic virus promoter). As viral coat protein is not produced, there is little concern over that impact, though the viral coat protein DNA is present, and can take part in recombination.
Furthermore, the transgenic plum was found to produce a novel small RNA molecule [2] associated with post-transcriptional gene silencing and virus resistance, and its safety to consumers has not been considered. Animals and humans may be exposed not just through consuming the plum, but also through breathing pollen or to the fruit juice through skin abrasions. This is particularly remiss, as interfering RNA species (RNAi) are now known to be ubiquitous, with many effects on biological functions [16, 17] (Life after the Central Dogma series; Subverting the genetic text, SiS 24). A small bacterial RNA was found to elicit RNA interference in mammals [18].
RNAi gene therapy, the injection of small RNAs to silence genes, touted as a ‘breakthrough’ in precision in 2002, was found to have significant off target effects in 2005. In May 2006, RNAi gene therapy was reported killing dozens upon dozens of mice [19, 20] (Gene therapy nightmare for mice, could humans be next? this issue). The effects were not sequence-specific. Out of 49 different sequences of RNA tested, 23 were lethal, killing the animals within 2 months. Another 13 sequences were “severely toxic” to the liver. Against this background, it is reasonable to ask if the small RNA in the transgenic plum is safe for humans and animals.
The USDA documents on the plums also do not appear to include any report on the impact of the transgenic plums on the mortality and behaviour of bees, which are the pollinators for plums. We should also ask whether the small resistance RNA produced in the transgenic plum is safe for bees.
The SAP report on PVCP-PIPS [11] provides poor guidance for risk assessment of the PPV-CP plum in yet another respect. There are well known post-transcriptional gene-silencing suppressors in the poty viruses related to PPV and in PPV itself; and the extent of homologous recombination between PPV and the PPV-CP transgene has not been adequately investigated. Post-transcriptional gene silencing and the stability of resistance of presumably the same transgenic plum had been studied earlier [21], and high levels of transgene mRNA were detected in the nucleus and low levels of transgene RNA in the cytoplasm. But the later reports [2, 5] stated that the inserted viral coat protein genes are methylated and show low levels of mRNA, indicating that transgene expression was unstable.
Transgenic C5 trees inoculated with virus appeared to show no infection during several years of virus exposure. However, the stability of PTGS has been questioned in studies showing that plum poxvirus silencing can easily be reversed through mutations in the small RNA targeting sequence, or by mutations that activate the virus’ suppressor of the host’s silencing [22, 23].
There is a clear need for fuller testing of the small silencing RNA from the transgenic plum for its effects on both plants and animals including bees and humans, and to consider fully the consequences of horizontal gene transfer and recombination and transgene instability.
Finally, there is the issue of transgene contamination. As bees are the pollinators for plums, it would not be surprising if the transgenic plum pollinates and contaminates non-transgenic varieties long distances away. The actual fruit of the plum will not be transgenic, but the pits of the fruit will be. USDA has deemed that accidentally GM-pollinated organic fruit is still organic, and has not commented on transgene pollution or on the patent-infringement issue, should the patent holders detect the transgene in non-transgenic plums, organic or otherwise. US has about 80 percent of the world prune export market and countries which look for transgenes in prune flesh will not find transgenes, but if whole prunes are measured the seed transgenes will show up.
The comment on organic plums and transgenic plums from the USDA petition is as follows:
“The presence of a detectable residue of a product of excluded methods alone does not necessarily constitute a violation of the National Organic Standards. The unintentional presence of the products of excluded methods will not affect the status of an organic product or operation when the operation has not used excluded methods and has taken reasonable steps to avoid contact with the products of excluded methods as detailed in their approved organic system plan. Organic certification of a production or handling operation is a process claim, not a product claim.
“It is not likely that organic farmers, or other farmers who choose not to plant transgenic varieties or sell transgenic plum, will be significantly impacted by the expected commercial use of this product since: (a) nontransgenic plum will likely still be sold and will be readily available to those who wish to plant it; (b) plum trees propagated by grafting and growers purchasing bud wood or grafted plants will know that this product is transgenic because it will be marketed as plum pox virus resistant plum. Additionally, decreasing the overall incidence of plum pox in conventional orchards may lower the likelihood of an organic orchard becoming infected.”
We would be very surprised if growers purchasing plants marketed as “plum pox virus resistant plum” would automatically know that the plants are transgenic. It is a blatant attempt on the part of the regulator to avoid labelling and hence mislead the public.
US organic plum producers may feel protected by the USDA position. But the export market may look at whole plums, not just the flesh of the plums.
Article first published 11/07/06
Got something to say about this page? Comment