The detection method for Syngenta's illegal GM maize is flawed; there must now be a full disclosure of information and access to reference material for retrospective risk assessment and risk management. Dr. Mae-Wan Ho and Prof. Joe Cummins.
Swiss biotech firm Syngenta had accidentally sold illegal GM maize Bt10 in the US for the past four years, resulting in about 133 million kilograms of the maize making its way into food and feed.
The news broke on 22 March 2005 in the science journal Nature ("Syngenta's GM maize scandals", SiS 26), although Syngenta had entered into talks with the US government since December 2004.
Under pressure from public protests across the world, the US government fined Syngenta a derisory US$375 000 (euro 270 000) for the mishap. And on 18 April, the European Commission imposed an emergency measure to ban certain GM maize imports from the US unless they are accompanied by an original analytical report issued by an accredited laboratory demonstrating that the product does not contain Bt10 ("Europe acts swiftly to keep out unapproved GM maize", SiS26).
Scarcely a week later, the EU authorities announced that Syngenta had presented a detection test for Bt10, which was already validated by the EU authorities.
The validation report [2] from the Joint Research Centre, also Europe's Community Reference Laboratory (CRL) for GM Food and Feed, said it carried out an in-house validation of the event-specific detection method "proposed by GeneScan on Bt10 maize developed by Syngenta Crop Protection AG."
Syngenta provided the DNA samples (genomic DNA extracted from the Bt10 maize line and from a control maize line), and GeneScan provided the event-specific detection method based on a qualitative polymerase chain reaction (PCR) assay.
So who, or what is GeneScan? GeneScan advertises itself on its website as "the world market leader in the field of molecular biological testing for Genetically Modified Organisms (GMOs) in food, feed and agricultural raw materials."
The GeneScan website has a link to a page on Syngenta's website, which advertises the "European Union Bt10 Detection Method" [3] as a "validated detection methodology that has been thoroughly tested for accuracy, reliability and sensitivity" using authentic samples to ensure actual targeted material is detected reliably when present. The method is designed, it says, to exclude "false positives" in the hands of "highly qualified scientific personnel with specific experience with the protocol", working under "exemplary laboratory practice and standard operation procedures (SOPS) from an …accredited lab", with "provisions for retesting false positives".
The same Syngenta page advises us that GeneScan is "the only private service laboratory that fulfils the elements listed above for Bt10 testing", and the fact that the EU Joint Research Centre has certified the GeneScan method on April 22, 2005 as "the only EU official method for Bt10 detection." Following that, yet again, the admonition to guard against "false positives" is repeated.
In contrast, there's not a word said about false negatives, which as every molecular geneticist knows, is also a problem with the PCR detection method, particularly if the GM insert is unstable, and prone to deletions and rearrangements, as revealed in recent analyses by European government laboratories ("Transgenic lines proven unstable", SiS20; "Unstable transgenic lines illegal", SiS21).
This three-way mutual reinforcement between Europe's Joint Research Centre (the European Commission's official laboratory), Syngenta and GeneScan seems just a bit too cosy to be reassuring. What's more, they have jointly declared a monopoly on the detection method, ruling out all others that could give "false positives". It is a case of the poacher turned gamekeeper with the help of the governor.
The validation report issued by the Joint Research Centre (JRC) goes on to state [2], "The results of the JRC validation demonstrated that the method reliably detects an amplification product specific for Bt10 maize, and therefore allows discriminating event Bt10 from other GM-events in maize lines. The sensitivity of the method is below 0.1%….
"The method is therefore considered by the CRL as fit for the purpose of Bt10 detection and it is the only accepted to certify the presence of Bt10 in maize commodities in accordance with the Commission Decision 1005/317/EC). (emphasis added)
In fact, the method amplifies and detects a small 130base pair fragment of Bt10 DNA, said to be specific for Bt10. It is not stated which gene fragment from Bt10 is being amplified. A strict protocol is laid out in detail. The Bt10 and wild type DNA supplied by Syngenta were analysed along with other reference and non-reference material contained in the JRC's Community Reference Laboratory.
The 130 bp band was indeed specifically amplified only in Bt10. But unfortunately, bigger bands were amplified and detected in other GM maize lines, and even in the wild-type maize DNA supplied by Syngenta. Strangely enough, these higher molecular weight bands were absent from the Bt10 DNA from Syngenta.
The origins of the "unspecific amplicons" (amplified DNA) were not investigated further, but effectively dismissed with the remark, "This suggests that the method can be further optimised." Consequently, only the 130bp amplicon is regarded as a definite positive.
The conclusion of the validation report states that the method is "fit for its intended purpose", with the qualification [3], "However, at this stage of testing, the method produces a higher molecular-weight multi-band pattern in GM and non-GM maize which requires additional efforts in its optimisation."
Still further qualifications are contained in a later report [4] on the detection method: "The analyst shall be aware that other validation experiments indicated that the method might perform less reliably at annealing temperatures higher than specified in the protocol. Moreover, in some incidents unspecific amplification was observed with PCR profiles that used high numbers of cycles than specified in the protocol. Time constraints did not permit to rectify these concerns…"
As mentioned earlier, fragmentation or rearrangements of the GM insert can change the size of the amplicon, or otherwise fail to give the specific amplicon. Consequently, unless fragmentation or rearrangement of the Bt10 GM insert can be ruled out, it is not legitimate to conclude that amplicons of other sizes are "false positives".
Syngenta's reports sent to the US Environment Protection Agency earlier this year have been leaked to ISIS.
The first report dated 28 January 2005 [5] is intended to present the DNA sequence of Bt10 compared with Bt11, the GM maize line that Bt10 had contaminated by accident. The Bt10 insert was mapped to chromosome 1 of the maize genome, while Bt11 insert had been mapped to chromosome 8. This alone will indicate that Bt10 is completely different from Bt11. In addition, there were three nucleotide changes in Bt10 compared with Bt11: two in an unspecified sequence contained within the Bt10 insert (unspecified sequence 1 in Figure 1 below), and one located in the nos terminator associated with the crylAb gene. No nucleotide changes were identified in any of the coding sequences and promoters within the Bt10 insert.
However, the map of the Bt10 insert presented can only be partial, as it did not include the ampicillin antibiotic resistance marker gene, unless that marker gene has inserted elsewhere in the genome. The map presented also contained at least three unspecified, unknown sequences (Fig. 1).
Unspecified sequence 1 (>1000 bp)-p35S (516pb)-IVS6 maize adh1S (477bp)-crylAb(syn) (1848bp)-tnos (267bp)-Unspecified sequence 2 (~400bp)-p35S(422bp)-IVS2 maize adh1S (180bp)-pat (522bp)-tnos (259 bp)-unspecified sequence 3 (~160bp)
Figure 1. Map of Bt11 from Syngenta's report to US EPA
The second report from Syngenta to the EPA is of a study comparing the transgenic proteins expressed in Bt10 compared with those in Bt11 [6]. The proteins were extracted from leaves of the plants, and subjected to western blot analyses, a technique dependent on staining the protein bands with specific antibodies after separating them by migration in an electric field through a gel matrix.
This report claims that the analyses "revealed similar dominant immunoreactive bands" in both Bt11 and Bt10 corresponding to the predicted Cry1Ab protein (for insect resistance) and phosphinothricin acetyltransferase (PAT) (for tolerance to the herbicide glufosinate ammonium) of about 69 000 and 22 000 daltons respectively.
However, the photographs of the western blots contained in the report tell a different story. Bt11 showed a series of bands at 46 000, 63 000 and 52 000 daltons (in order of strength of staining) besides the dominant 63 000 daltons band, whereas Bt 10 only had the 63 000 daltons fragment besides the main predicted band. The PAT protein bands in Bt10 and Bt 11 were also different from each other and from the purified standard, with many high molecular weight bands reacting to the antibody.
Neither report contains information on the breeding history of the GM maize lines analysed, such as the number of generations since the transformation event; nor data from appropriate reference material. These are sure signs of sloppy science.
The detection method for Bt10 is flawed by the admission of the European authorities. The identity of the 130 bp amplicon, supposed to be specific for Bt10, is not made explicit. The molecular data supplied to the US EPA are incomplete. It is impossible to judge if the detection method is adequate in the absence of full molecular data including those from reference material proving that Bt10 had remained genetically stable since it was first unintentionally released.
Bt11 had already been exposed to be unstable, and to be contaminated with another Syngenta maize Bt176, implicated in the death of dairy cows in Hesse Germany ("Cows ate GM maize and died", SiS 21).
Syngenta has admitted that Bt10, as distinct from Bt11, contains an ampicillin resistance marker gene, which, according to an Opinion issued by the Scientific Panel on Genetically Modified Organisms of the European Food Safety Authority in 2004,
"should not be present in GM plants to be placed on the market". No official information has been forthcoming regarding the ampicillin resistance marker gene in Bt10, nor any attempt to ascertain whether the marker gene has contaminated other maize varieties, GM or otherwise.
As Bt10 has already entered the market and the human food chain, it must go through retrospectively the risk assessment process that would have been applied to a GM product approved for market. This is also essential for effective post-release risk management.
At the very least, Syngenta must be required to provide the following:
Furthermore, regulatory authorities on both sides of the Atlantic must make public all information on Bt10 that they have received from Syngenta or other sources.
Please circulate this report widely and send it to your elected representatives.
Article first published 18/05/05
Got something to say about this page? Comment