1.Are Bt10 and Bt11 the same? - Greenpeace Australia
2.Critique of Syngenta's documents - Dr Jack Heinemann
While a shipment of Syngenta's rogue Bt10 corn has just been impounded in the EU, in the US and elsewhere it is being claimed that Bt10 is for all intents and purposes the same as an approved variety of Syngenta corn - Bt11 - and it is therefore safe for human consumption.
It is fast becoming clear that this is not so.
"The Syngenta documents you have provided indicate that there are additional and possibly substantial differences between BT10 and BT11." - Dr Jack Heinemann, Institute of Gene Ecology, University of Canterbury
1.Are Bt10 and Bt11 the same?
In mid-May, Food Standards Australia New Zealand (FSANZ), responding to increasing public pressure and a request from the Geneethics Network, released several documents that purport to support the FSANZ view that Bt10 is virtually identical to Bt11.
FSANZ has argued that "the two varieties have been modified in the same way and produce the same novel proteins. The presence of a non-functional antibiotic resistance marker gene (BLA) in Bt-10 corn, that is not present in Bt-11, has no impact on the safety of food produced from Bt-10 corn."
FSANZ has argued that because Bt10 is for all intents and purposes the same as Bt11 and Bt11 has been deemed safe for human consumption, then FSANZ is justified in taking no steps to remove potential Bt10 products from Australian supermarket shelves or to prevent possible continuing imports of Bt10 corn products.
In doing so, FSANZ has circumvented and ignored their own processes for assessing GE foods and dealing with prohibited imports such as this.
And now, even the claims of safety are looking less secure.
[See] the Syngenta documents released by FSANZ and the critique of those documents prepared by Dr Jack Heineman of the NZ Institute of Gene Ecology at the University of Canterbury in NZ. [below]
These Syngenta documents do not represent all the data available to FSANZ, only what they have chosen to release in support of their decision to take no action in relation to Bt10.
The Heinemann critique makes it clear that based on these documents some differences between Bt10 and Bt11 can be established.
"The Syngenta documents you have provided indicate that there are additional and possibly substantial differences between BT10 and BT11."
Further, claims relating to the similarity of Bt10 and Bt11 cannot be ascertained from the materials released.
"The report SSB-112-05 indicates that there were differences in the profiles of PAT and Cry1Ab proteins and thus there may be other undetected differences."
The most concerning analysis in the Heinemann document is that claims of the safety of Bt10 may need reassessing based on the Syngenta documents released. Heinemann notes:
"The PAT levels in BT10 appear to be much higher than in BT11. This is not readily explained by the overall DNA sequence similarity between the BT10 and BT11 events. Moreover, the differences in concentration could indicate a higher overall stability of the PAT protein, and possibly Cry1AB protein (if it is true that the lower molecular weight bands are indeed breakdown products in BT11), produced by BT10. If indeed these proteins are inherently more stable, or delivered to consumers at higher concentrations, their potential to be allergens may be underestimated by studies using BT11 as the source."
These documents are now available on the Greenpeace Australia website. Follow the links from the feature at: http://www.greenpeace.org.au/features/features_details.html?site_id=45&news_id=1672
2.Critique of Syngenta's documents - Dr Jack Heinemann
New Zealand Institute of Gene Ecology
Tel: +64 3 364 2500, Fax: + 64 3 364 2590
Re: Syngenta Biotechnology Report SSB-104-05 and SSB-112-05 on BT10
Thank you for forwarding these two reports to us. I have read the reports and will attempt to answer the question you presented in your email of 13 May, 2005.
By way of background and outside the context of the documents you sent, it has been reported that the BT10 line differs from the BT11 in that BT10 retains the gene that confers resistance to the antibiotic ampicillin on some bacteria [Nature Biotechnology v23:514 (2005)]. I can find no reference to the resistance gene in the reports you sent. Since the insert in BT10 was sequenced for purposes of establishing its similarity to BT11, I assume that the antibiotic resistance gene has integrated elsewhere, or at least outside of the sequenced region. It is impossible for me to determine the importance of that integration event without knowing where it is, what genes it may have integrated into or what novel hybrid (or fusion) proteins might emerge from what is called a read-through into the ampicillin-resistance gene.
In addition, it is a common phenomenon for transgene constructs to integrate in multiple
places in the genome, and for very small parts of the construct to integrate independently of full sized versions (references for this statement can be found in our submission on A549 which is available on our website). Without knowing if, or how well, the genome of BT10 was inspected for potential secondary insertions of full or partial composition, I also cannot determine the relative equivalence of BT10 and BT11. Since the DNA for these studies was sourced from plants that were derived from a series of crosses to the original BT10 line, it is impossible to say how their genomes would represent the genomes of the plants (if indeed those lines are different) that have mixed with commercial BT11 lines, and whether by chance they have retained or lost the chromosomes that may carry other insertions.
Impacts arising from uncharacterised insertions cannot be predicted from characterised insertions. Impacts from other insertions might be seen if BT10 had been subjected to a "substantial equivalence" test. While such tests do not prove equivalence, they can provide evidence of alterations in genes or genome function that result in changes to the
overall normal composition of the organism. The documents you provided do not include such an analysis, nor do they discuss other "product-oriented" tests, such as the results of feeding studies. Again, without such information, I could not comment on the equivalence of BT10 and BT11.
The reports do make clear several other differences between BT10 and BT11. First, the characterised insert in BT10 resides in chromosome 1 whereas in BT11 the insert is in chromosome 8. The context of the insert would obviously be different, as would be the genes that may be affected directly by the insertion. Second, there are three recorded basechanges in the sequenced part of the insert in BT10 that is common to BT11. These three nucleotide changes were the essence of your question to me, but my answer will include other context issues too.
The context of the insertion is relevant for several reasons. As indicated above, the integration may have effects on adjacent genes by, for example, introducing a change in chromatin structure or inducing a change in methylation (which often is initiated de novo from duplicated DNA sequences that may arise from secondary but unknown inserts).
While it is possible that the insert may reside in a region of no known function, (cis-acting) regulatory DNA sequences, for example those with weak promoter or enhancer activity, are still difficult to identify and we cannot exclude the possibility that one or more such sequences were changed as a result of the insertion(s) in BT10. Furthermore, read-through transcription””initiated somewhere in the insert and ending outside it, or initiated in adjacent regions and ending in the insert””may be the source of novel RNA and proteins. Abortive transcription from read-through might, for example, produce novel short and double-stranded (ds)RNA molecules. A risk factor emerging from the production of novel dsRNA is the potential to induce gene silencing either locally or on other genes (again, a reference list is provided in our submission on A549).
The collection of DNA elements that have been used for the construction of the BT10 event have both known functions and probably unknown functions; the latter can only be identified by analyses that go well beyond DNA sequencing. It has recently been demonstrated, for example, that DNA sequences of the nos terminator, used in two places in the BT10 event (and perhaps distributed elsewhere in the genome), may be cis-acting alternative splice signals (see references to submission on A549). As you are aware, the DNA sequence is a template for the transcription of an RNA molecule. Less commonly appreciated is that the RNA molecule may undergo a large series of transformations, including splicing to remove introns, but also alternative splicing, that results in families of different RNA molecules all derived from the same original source. These families do not necessarily give rise to the same proteins or proteins with similar functions. While the nos terminator is not derived from an organism like corn, it may be processed by the enzymes that mediate splicing and generate novel RNA molecules in the process. It is in my view reasonable to query the effect of this cryptic splicing activity on the transcriptome of the BT10 lines; it is even more important given the nucleotide change reported in nos in BT10.
The other two reported DNA sequence changes are adjacent to the 35S promoter used to initiate transcription of the cry1Ab(syn) gene. This is in a region called an "intervening sequence", presumably so-called because it has not been demonstrated to be transcriptionally active. I also presume that the assignment was made in the donor organism, not in the corn plant. The level of demonstration is not made clear in the report so it remains formally possible that the assignment of "intervening sequence" is based on a bioinformatic analysis and not empirical testing. I would have low confidence in this assignment if the conclusion were drawn only from bioinformatic analysis.
Regardless of how it was determined that the DNA adjacent to the promoter was an “intervening sequence”, since the DNA was sequenced it is reasonable to assume that it is part of the construct and new to that position in chromosome 1. Its potential to contribute novel cis-acting regulatory influences has not been tested in the set of experiments described in these reports and should not be dismissed without test, both because it is adjacent to presumably different genes in BT10 than in BT11, and because it is adjacent to presumably very different DNA sequences in chromosome 1 than in chromosome 8. It is clearly documented that introns can influence promoter activity and RNA processing reactions such as splicing and RNA editing, making premature the conclusion that this "intervening sequence" is functionally inert without a test.
The Syngenta documents you have provided indicate that there are additional and possibly substantial differences between BT10 and BT11. If the position of the insert (chromosome 1 vs. 8) and effects of other contextual elements (such as the altered nos
and intervening sequence) were of no consequence, I would expect that the profile of purified Cry1Ab and PAT proteins from BT11 and BT10 would be identical. If they are not, then we should, in my opinion, rule out other changes. The report SSB-112-05 indicates that there were differences in the profiles of PAT and Cry1Ab proteins and thus there may be other undetected differences.
Interestingly, three bands are visible in the Western blot of Cry1Ab beyond the band that migrates at an apparent molecular weight of 68,900 Da. Two of these are acknowledged in the conclusion of the report, those migrating at an apparent molecular weight of 46 and the 52 kDa. Another unacknowledged band appears at approximately 67 kDa. The two identified bands of lower molecular weight are said to be “minor breakdown fragments”, presumably of Cry1Ab protein. Nothing in the report confirms this identification and it is well within the abilities of protein science to make such a confirmation. It is also possible that they are proteins with a common epitope to Cry1Ab but derived from a different mRNA (due to alternative splicing or read-through). It is a concern that the relative concentration of these alleged breakdown products is very different in BT11 and BT10.
The concentration of Cry1AB (69 kDa species) and the ~67 kDa species is nearly the same in both plants, but these bands are barely visible at the 46 and 52 kDa region of the blot for protein preparations isolated from BT10. (In fact, when I printed the report for my use I could not see these bands. They became only just visible when they were printed using a different printer!) I would expect these profiles to be identical unless the proteins detected were not breakdown products at all, but legitimately synthesised proteins of different mRNA transcripts. I also do not know whether their characterisation as breakdown products was ever challenged by regulatory authorities evaluating BT11. This new work by Syngenta, along with the very recent observation of transcription across the nos terminator and processing of transcripts that include the nos region, may provide reason to reconsider the safety analysis of BT11.
Different inconsistencies are apparent in the PAT Western blot. The PAT levels in BT10 appear to be much higher than in BT11. This is not readily explained by the overall DNA sequence similarity between the BT10 and BT11 events. Moreover, the differences in concentration could indicate a higher overall stability of the PAT protein, and possibly Cry1AB protein (if it is true that the lower molecular weight bands are indeed breakdown products in BT11), produced by BT10. If indeed these proteins are inherently more stable, or delivered to consumers at higher concentrations, their potential to be allergens may be underestimated by studies using BT11 as the source.
In summary, it is not possible for me to assess the importance of the three reported base changes in the characterised BT10 event, because the importance of DNA sequences is in part derived from their context and contribution to higher-order events on the RNA and protein level. While a difference of three nucleotides may seem small, we know that it may not be unimportant. Clearly the events are in different contexts (by being on different chromosomes) and the Western blots suggest, but do not prove (since they were semi-quantitative), that there are significant differences in expression profiles between BT10 and BT11.
The New Zealand Institute of Gene Ecology is a public research centre that emphasises responsiveness to the research questions from the public, particularly those who do not have direct access to the resources to conduct their own research. We have conducted this analysis free-of-charge as part of the University’s role as critic and conscience of society.
Assoc. Prof. Jack Heinemann, PhD
18 May 2005
Cry1Ab Syngenta recombinant allele of a gene sourced from Bacillus thuringiensis.
Confers resistance to certain insects.
PAT phophinothricin acetytransferase (herbicide tolerance).