A report from an activist group called Safe Food Foundation (SFF) came out last fall that caused a minor stir upon it’s first release. They claimed that they had unearthed an issue with GMO wheat being studied by the Australian CSIRO researchers. The wheat under investigation has shown to provide improvements in digestive health in animal studies and could potentially lower the glycemic index of foods. SFF threw a press conference, did a YouTube video, and managed to get some press about it. Here’s a New Zealand newspaper that picked up the claims. But as the story unfolded, it quickly became apparent that their claims were wrong and irresponsible.
There’s an astonishing amount of misinformation propagated around the internet about GMOs. Most of it is so clearly science fiction and conspiracy theory that it’s just laughable. Sometimes, though, the information comes from what seems like a trustworthy source, such as a scientific journal. Even if it is published in a respectable journal, sometimes the statistics and results do not withstand scrutiny. Science is a process and claims made in publications need to be evaluated over time.
Another source of confusion can be the “gray literature“. Gray literature consists of publications such as reports that have generally not gone through the peer-review process that scientific studies generally do. Some gray literature is useful and valuable, and quite neutral, based on a balanced panel of participants with appropriate skills. Sometimes, though, there are reports that come out that are definitely not neutral, and represent merely the cherry-picked and distorted views of an industry or activist viewpoint, sometimes both. The Safe Food Foundation report on CSIRO’s wheat was one such example of gray literature.
As reported by the press: The scientists claim
* SiRNA, a form of ribonucleic acid, like DNA, could transfer to humans through food when produced in GM wheat.
* When eaten, the siRNA engineered to suppress the wheat-starch branching enzyme (SBE) would also silence the human-branching enzyme which produces energy-storing glycogen (GBE).
* This “unbranched” glycogen would have low solubility in human cells and could create build-up in the tissues of the body, especially in the heart and liver.
* This could lead to the disease Glycogen Storage Disease IV, resulting in an enlarged liver, cirrhosis of the liver, and failure to thrive.
The claim was that the sequence used in the CSIRO wheat would destroy the activity of the human GBE gene, leading to the same outcome as this disease that kills children by the age of five. Sadly, though, it was one of those cases of a biased report that would not have passed any legitimate peer-review. And although the press did bite because of the shocking claims, mostly the claims only made the rounds of the usual suspects who will reprint any shocking claim, without any ability to assess the validity.
Here at Biofortified we immediately saw what was wrong with this report. It was based on copious amounts of incorrect sequence data, and on sleight-of-hand with bioinformatics tools. I prepared a long technical post about the flaws in the claims of the Safe Food Foundation’s team of hired guns Jack Heinemann and Judy Carman, but as an obscure report that was so obviously wrong, we decided it wasn’t worth drawing any more attention to it.
It recently resurrected, though, at an unusual spot. Orac at Respectful Insolence actually took it on: Oh, no! GMOs are going to kill your babies and permanently change your gene expression! He does a good job of framing the issue and explaining the technology. He astutely writes about this report:
Of course, the problem with Dr. Heinemann’s highly speculative analysis is that he didn’t know the actual siRNA sequences that were going to be used. Without that information his analysis was pretty pointless. At the very best, it was highly speculative. At the worst, it was ideologically and politically motivated.
I know my way around bioinformatics – in fact I teach it to others. So let’s take a brief trip through the DNA sequences in question to see what was really wrong about the Safe Food Foundation report. The concern they express, resurrected here, is that double-stranded RNA (dsRNA) produced by RNAi silencing in GE crops could silence our own genes if we eat them. The mechanism is complicated, but in short, before there can be any silencing there needs to be a sequence match between the dsRNA produced by these plants and the DNA of our own genes. Heinemann claimed that he found these sequence matches – but as I will demonstrate he didn’t find anything of the sort.
The claims Heinemann made were based on a sequence that was 25,187 bases long. In no way that was the appropriate sequence to use for the analysis. It was like throwing 25,000 strands of spaghetti at the wall to see what sticks. There were going to be false positives. And one of the pasta pieces that stuck was to the GBE human gene. This was the basis for the claims made by Judy Carman in her SFF report that provided the dire warning about how this wheat would kill your babies.
Within the large record, only a small portion represents the SBE1 gene. See the long image on the right. And even within the actual gene part of that record, only a fraction of that sequence would be in the constructs used. The key point to be made about this was that their analysis was pointless and wrong.
I learned from Orac’s piece that there had been an update to the analysis, though. Heinemann had apparently realized that the sequence he used was probably incorrect. (Yes, we here at Biofortified had spotted that immediately.) There’s a re-analysis with a different sequence that you can now examine.
The new sequence? Oh, it’s different, yeah. It’s found within this GenBank record. But it’s not even the whole record. It’s only bases 96-635. That’s right–540 bases, not 25 thousand. Mm hmm. In case you are curious, 25,187 / 540 = more than 45x too much sequence. That’s a lot of pasta.
So there’s less spaghetti now. But again, Jack is able to deliver some sticking to the wall. However, all of the matches he highlights are either to introns – which would not matter for the mechanism of action that is the issue here. Or they are in the genome desert areas, thousands of bases away from anything that appears to be a gene.
Result: Take a deep breath. The GMO wheat that forms the basis of this claim will not kill your children or permanently alter your genome.
More interestingly, I decided to check back and see if the human GBE gene is still sticking with this sequence switcheroo. The main fear-based claim was that the GBE gene would be suppressed and children with a disorder of this can die by the age of 5. So let’s see–let’s do an analysis of this new sequence and compare it to GBE. If you perform a BLASTn at NCBI with these two sequences and default settings as Jack says he does, what do we get?
The result? No significant similarity found.
That’s right. The entire foundation of the fear–the match to GBE–is gone. Evaporated. Vanished. Nada. Zip.
And yet people who don’t understand the data will continue to make this claim. They will continue to be wrong.
This is really a case of shouting “fire” in a crowded theater. Fires are something to watch out for. And it’s important to have professionals who assess building codes evaluate the theater, as we do for GMO foods. But causing a stampede for the exits also has consequences and can be harmful in its own way. Sadly there’s a lot of this on this topic.
What can you do?
- Don’t base your decisions on scary graphics. They may not be right.
- Be aware of the source of the information. And understand the difference between the peer-reviewed process and the gray literature.
- When you do see claims, check them out. Or find a person qualified in the field to help you to understand the claims. Conspiracy theorists may not be a good place to find validation.
If this was a journal publication, I could write to the editor and ask for corrections or even a retraction. But that’s another problem with the gray literature. There’s no mechanism to do that. I checked with Retraction Watch to see if there was anything they knew of, but there isn’t any kind of a registry or anything. We can only hope that people will Google for facts. Unfortunately they will be swamped with the misinformation instead. At least you have the knowledge you need now. Do share it.
Regina, A. (2006). High-amylose wheat generated by RNA interference improves indices of large-bowel health in rats Proceedings of the National Academy of Sciences, 103 (10), 3546-3551 DOI: 10.1073/pnas.0510737103