Wednesday, January 11, 2012

Ari LeVaux May Have a Point

Edit: 2/4/2012

Are you still confused about the biology of this whole thing? I don't know how anyone could not be as the articles written about it conflict with one another. If you are not a biologist, you will be further confused if you read the post below, as it was directed an audience with more background. I tried to do a more simple explanation here, where I attempted to explain the science behind some of the biology in contention. It should only require that you know what DNA, RNA, and a protein are. If you read that first you will understand this better. Feel free to ask questions if you want. No one reads this blog (except you! Welcome!), so I don't have too much going on, blog-wise. As for me, I am a biologist who will soon be graduating with a PhD. I study DNA and RNA, though not GM food specifically.

Here is my original post:

Ari LeVaux wrote a post arguing that new information suggests a need for greater testing of genetically modified (GM) food before it is allowed into our food supply.

This argument is based on the premise that small RNAs that are in the food we eat can act on our physiology. Small RNAs (miRNAs and siRNAs introduced by RNAi) bind to larger mRNAs which causes the mRNA to get chopped up, and lowers the expression of individual genes, to prevent them from acting.

I think two lines of evidence to support this idea.

The first is a study from a Chinese group that Ari cited. This study found that small RNAs in rice are present in human blood and that they can potentially change the LDL levels of people who ingest them. This study is making pretty remarkable claims and their findings will be need to be replicated in other labs.

The second line of evidence comes from Monstanto themselves. Monstanto uses the idea of small RNAs being ingested and acting on the host to kill bugs. They integrate small RNAs into corn that, when eaten, will disrupt the gene expression pathways of the bugs and kill them.

Ari argues that the idea that small RNAs from food may be affecting human health renders obsolete industry's claim that testing of GM foods is not necessary.

The industry objection to performing human safety trials on GM foods is in part based on the substantial equivalency argument. Substantial equivalency argues that these genetic modifications introduced into the the genome of GM food are safe based on the idea that the biological active agents that they are introducing into the food supply have been eaten before and therefore have a history of safe use. When this was based on central dogma this was true - the new proteins introduced had a history of safe use.

But techniques that are used in genetic food modification also introduce small RNAs into the genome of food we eat. The first genetic modification approved for human use was a knockdown of a gene's expression that was used to prolong the shelf life of a tomato which inadvertently introduced small RNAs into the tomato. If you are introducing novel small RNAs into a GM food supply you cannot logically argue a history of safe use because the small RNAs themselves have no history of use at all. You are introducing a new small RNA into the food supply that has never been ingested by a person before. But RNAs as a class are not currently tested for biological effects.

How great is this potential for harm? I am speculating because I do not have the actual RNA sequences, but I think it's super small. The way small RNAs work is that the sequence of the small RNA has to be an exact or almost exact match for the sequence of the gene that it is knocking down. The chances that there will be an off target hit in great enough quantity to do harm seems remote.

Can GM food makers mitigate the risk cheaply, without testing the food? Yes. In cases where there is not a history of safe use the GM food makers are permitted to use other approaches to establish that the food is safe, meaning that you can introduce novel chemicals into the food supply by GM but you have a greater burden to prove they are safe. There are computational methods that can be used to demonstrate the small RNAs are unlikely to be hitting any human genes. For example, I just checked the gene that was knocked down in the tomato example and it does not have close matches in the human genome. This does not mean it could not have off target effects - I have to run it through some more programs for that. But this risk is low. However, the risk cannot be mitigated completely with computation methods because we are still working on ways to more accurately predict which small RNAs will bind to which genes, especially across diverse populations.

Is this a reason to clean the GM food out of your cupboards? No! GM food has the potential dramatically increase our food supply, make our food more nutritious and stabilize the food supply. I am actually a big fan of the potential of GM food and usually find that the arguments of the wealthy worried are thin.

Is this a fair question for a food writer to ask? Yes, of course it is! It's an awesome question. The burden to prove that these foods are safe falls on Monsanto, not on the consumer. And the new research does raise legitimate questions Monsanto needs to address if they are going to be putting these modifications into our food supply without even the minimal requirement that GM food is labelled.

Even though I expect that the danger from small RNAs is small, if the finding that small RNAs from food are acting in humans holds up it will certainly be used to bolster the arguments against GM food. In 1994, when GM small RNAs first hit the food supply, a comprehensive analysis of off target hits against the human genome certainly could not have been performed because the human genome draft was not done yet. And when Monsanto introduced a small RNA technique a few years later that used small RNA to act on the host it was a big surprise to a lot of people outside Monsanto that they could do this. Now, if ingested small RNAs are acting in humans as well, it certainly supports the argument that it is arrogant of the GM food companies to assume that they know everything about the biology they are using to modify the food.

I don't know how valid this argument is myself but I expect I'll be reading it other places.


  1. I'll also add my own views on the point of GM food labeling: GM food should be labeled. Food is more than the sum of its chemical constituents, and it's completely reasonable to think eating food from a lab is icky. No one says, "Mmmmmmm, modified...." For this reason, Monsanto and the other companies that make GM food have a marketing problem aside from the perceived risk. It is not the government's job to solve Monsanto's marketing problem for them. It's the government's job to do what the people want, and if consumers want their food labelled then it should be labelled.

    And it's not necessary to invoke any science, good bad or otherwise, supporting a possibility of harm to justify that.

  2. This comment has been removed by the author.

  3. There is also a post on Dr. David Tribe's site that discusses the history of safe use of small RNAs in the food supply:
    I would also like to point out that the digestive systems of insects and mammals are very different, and is has been pretty easy to demonstrate that small RNAs can be taken up in insects, and more difficult to demonstrate in mammals. I follow this pretty closely for honeybees, where small RNAs in sugar syrup can not only silence viruses in the bees, but also a gene in a parasite that feeds on the bees. There are risks and issues with small RNAs being engineered into crops, but I think Ari Levaux over-blew the implications of this paper. Interestingly enough, he said that he still eats rice, which means he's not actually concerned about a potentially medically-relevant micro RNA that was discovered in the paper, that has unknown health consequences.
    He is apparently going to edit the column, I'm interested to see how he changes it given the response he has gotten.
    *Note: edited version of previous comment - will use preview function next time!

  4. -more difficult to demonstrate in mammals

    I think it's a question of a threshold of proof, and whose responsibility it is to do the proving.

    The Zhang study demonstrated that the small RNAs were taken up during digestion to a threshold that was good enough for their reviewers to publish it.

    So do we have to do more to prove to Monsanto that they are active or is this paper enough to throw the ball into Monsanto's court to have to prove to us that they aren't? People had issues with the paper, but I don't think something like this needs to be proven with 100% certainty if it is questioning the safety of something that is in the food supply, unlabelled.

    I don't know that Ari need to prove that the small RNAs are dangerous. He just needs to prove that the small RNAs introduced are potentially biologically active agents that weren't adequately considered when these foods were approved. That is, that the approval process was inadequate in the light of new information.

  5. Also, Karl, not to pick on you because I've seen this everywhere:

    I don't think it's logical or good science to say that we eat miRNAs in food that is safe, so therefore all miRNAs are safe. That is the basic premise of the Monsanto History of Safe Use paper.

    That is only safe IF they are safe as a general class because they don't survive digestion in tact in great enough quantities to change human physiology.

    If that is not true (as the Zhang paper which needs to be replicated implies) then you have to treat each miRNA as a separate chemical entity.

    Just as you treat each protein as a separate chemical entity. e.g. you can't integrate botulism toxin into GM food because it is a protein and proteins are present in high quantities in safe food.

  6. The findings of the Chinese study were replicated by these guys: