A new study claims we should be less worried about GMOs because the sweet potato has been genetically engineered by nature herself

An article for is headlined, “Genetically modified crops? Nature got there first: The sweet potato has been genetically engineered by bacteria”. Reviewing a new study published in PNAS, the article explains:

“One of the most frequently mentioned issues with GMO foods is a vague concern about bringing genes from distantly related organisms into plants. But an international team of biologists has now found that this has occurred naturally in a major crop plant: the sweet potato. The strains of this crop that we currently cultivate had a set of genes inserted into their genomes by bacteria — the same bacteria used to create many genetically modified plants.”

NPR joins in with the hype: “The first genetically modified crop wasn't made by a megacorporation. Or a college scientist trying to design a more durable tomato. Nope. Nature did it — at least 8,000 years ago. Well, actually bacteria in the soil were the engineers. And the microbe's handiwork is present in sweet potatoes all around the world today.”

The bacterium in question is Agrobacterium. In genetic engineering, this same bacterium is used to ferry the GM gene (transgene) into the host plant. The study shows that the Agrobacterium genes got into the sweet potato by natural horizontal gene transfer (HGT). HGT means that genes are transferred by a method other than sexual reproduction (sometimes called vertical gene transfer).

The authors of the study express a hope in their published paper that "Our finding, that sweet potato is naturally transgenic while being a widely and traditionally consumed food crop, could affect the current consumer distrust of the safety of transgenic food crops” and “influence the public’s current perception that transgenic crops are ‘unnatural’”.

Is this justified? Has nature genetically engineered the sweet potato long before humans thought of the technology, and if so, should that reassure us about GMO safety? GMWatch asked scientists what they thought.

One thing quickly became clear: a huge amount of spin has been put on this paper.

That’s because far from reassuring us that GM is safe, the sweet potato discovery simply confirms what scientists have long known: that horizontal gene transfer takes place in nature as well as in genetic engineering.

But the crucial difference is that in nature, it takes place over evolutionary time, with potential negative consequences along the way that are completely unknown to the scientists who wrote the recent study and to everyone else. Long co-evolution keeps any harms localised and limited while human and animal consumers learn which plants are safe to eat and which are toxic or otherwise dangerous.

With GM, horizontal gene transfer is telescoped into a short time period into a large global acreage of crops without the benefit of co-evolution over long periods of time.

The added irony is that it is critics of GM who originally flagged up the whole issue of horizontal gene transfer (HGT), pointing out that it meant genes deliberately introduced by genetic engineering into one organism might move into another by HGT. This concern was dismissed by defenders of GM who, as Dr Michael Hansen notes (below), either "denied that HGT happened or claimed it was unimportant". Now they have adopted the reverse strategy and are busy hyping HGT as if it were far more common than it actually is and as if it showed genetic engineering to be "natural" and perfectly safe.
The scientists’ comments follow.

Michael Hansen, senior scientist, Consumers Union, said:

“This paper validates what GMO critics have said all along: that horizontal gene transfer (HGT) is a potential risk of GM and must be considered as part of the risk assessment – yet it misleadingly presents this fact as showing that GM technology is safe.  

“Historically, GMO seed companies have denied that HGT happened or claimed it was unimportant, since they were arguing against having to look for any unintended consequences due to the insertional mutagenesis associated with HGT.

“Indeed, I wrote the report, ‘Genetic Engineering is Not an Extension of Conventional Plant Breeding’ in January 2000 to draw attention to the risk of unintended consequences due to the insertional mutagenesis associated with HGT.

“The US FDA explicitly recognized this risk in 2001, when it proposed requiring companies to notify the government at least 120 days before commercializing a transgenic plant variety and named insertional mutagenesis as potential problem: ‘Because some rDNA-induced unintended changes are specific to a transformational event (e.g. those resulting from insertional mutagenesis), FDA believes that it needs to be provided with information about foods from all separate transformational events, even when the agency has been provided with information about foods from rDNA-modified plants with the same intended trait and has had no questions about such foods. In contrast, the agency does not believe that it needs to receive information about foods from plants derived through narrow crosses [in traditional breeding].’

“In other words, the US FDA admitted that there is a difference between GM and traditional breeding and that unintended consequences associated with insertional mutagenesis should be assessed. In spite of this, FDA is still following the 1992 policy (that there is no difference between GM and conventional breeding, a policy that came out of the White House Council of Competitiveness and was introduced at a BIO meeting by the then Vice-President Dan Quayle as a deregulatory initiative), rather than the 2001 policy.

“The notion that this natural engineering of sweet potatoes shows that GM technology is perfectly safe is false. Since we weren't around to document the early history of these sweet potatoes, we have no idea if they caused problems.

“Let's assume that the first ‘natural’ GM sweet potato, in addition to having some of the Agrobacterium DNA present, also, as an effect of insertional mutagenesis, caused a gene to be turned on that produced birth defects, sterility, or reduced fertility. As the further breeding occurred there would be variable levels of this particular toxin among sweet potatoes. People eating the sweet potatoes with high levels of the toxin would have fewer viable offspring, so the process of natural selection (the co-evolution of people and the food plants they are domesticating) would result in a shift toward decreasing the level of the toxin in sweet potatoes, due to the strong selection pressure against higher toxin levels.

“Since the Agrobacterium DNA has no direct link to the toxin, there would be no selection pressure to remove the Agrobacterium DNA. Thus it would persist in the modern traditional varieties developed from the initial naturally transformed sweet potato, enabling the team of scientists to arrive at their findings and write the paper under discussion. But far from reassuring us that genetic engineering is safe, all that can be concluded from the new paper is that the scientists have no idea what the history of the development of this sweet potato might be, or what effects it might have had on human or animal consumers during its evolution.

“Similar arguments have been raised regarding GM golden rice. Golden rice needs to be evaluated to see if levels of retinoic acid (a known teratogen), or any other potentially toxic retinoids, have increased in golden rice as an unintended effect. GMO proponents have argued that if high beta-carotene levels could lead to high retinoic acid levels and increased birth defects, why haven't we seen such problems with people eating carrots, or other foods high in beta-carotene? The answer is that humans have co-evolved with their food plants over time, so that if there had been varieties of carrots that caused such problems, there would be negative selection pressure against those traits so the toxin level would decline over time. With golden rice, there has been no co-evolution, so that's why it must be checked for potentially increased levels of potentially toxic retinoids.

“The scientists who wrote the sweet potato paper may not know the history of the HGT/insertional mutagenesis issue, and could be naively making the argument that natural HGT in the sweet potato means that crop genetic engineering is safe.”

Dr Michael Antoniou, a London-based molecular geneticist, said:

“Maybe Agrobacterium genes did insert into the sweet potato genome. But then the genetic alteration was selected for fitness, advantage, and crucially, food safety for humans and animals over evolutionary time. This does not happen with GM crops. And the discovery that Agrobacterium gene sequences have ended up in the sweet potatoes does not equate with the artificial combinations of DNA sequences that make up the GM gene units that are introduced into GM crops. So this natural event cannot be likened to GM technology.”  

Prof Jack Heinemann, University of Canterbury, New Zealand, said:

“This is an example of horizontal gene transfer (HGT). There is nothing unanticipated in this discovery because the process has been known to occur for decades and many other processes of HGT from bacteria to multicellular organisms have been demonstrated in the laboratory. So we’ve known for over a half century about HGT, even between kingdoms of organisms. Agrobacterium DNA in sweet potatoes isn’t ‘natural transgenics’; it's the outcome of a natural process.

“Ironically, especially in the 1990s many who were developing or selling GM crops attempted to minimise the frequency or effects of HGT. Indeed, they also tried to define it in ways that were so restrictive and unusual that it would be difficult to prove. When contradictions to this view inevitably appeared, it was reconstructed as evidence that HGT is common and therefore HGT from or to GM crops was no different than HGT from anything else. That somehow equated to it not being a new risk.

“The point is that when we move genes, we create organisms with no history of safe use and they should be tested for safety, consistent with international agreements. When people move genes into plants, we move constructs that we have pieced together from an average of 8 different species simultaneously. In my 25 years of work on HGT, I've seen no precedent for this kind of transfer so quickly. When HGT occurs in nature, nature has a chance to react, respond and adjust over many millennia to initially very small descendant populations. When we do it, nature is immediately bombarded by millions of hectares of new organisms in only a few years.

“Of course nature can also create organisms - by HGT or other means - that are capable of causing us harm. But that is no reason for us to do it unwittingly to ourselves. Nature can squash us with a rock from space, causing injuries indistinguishable from a car crash. This is not a reason to stop motor vehicle safety testing or recommend removing seat belts.”

Dr Ignacio Chapela, associate professor, University of California, Berkeley, said:

“There is nothing new here, and no surprises. We have known about this for almost forty years.

“What the authors of the new paper — and the reporters writing about it — claim as a surprise is based on the existence of a sequence incorporated over evolutionary time into the genome of a plant. This introgression implicated long processes of trial-and-error in a complex context which cannot be reproduced in the genetic engineering laboratory. The ‘thing’ resulting in the end may look similar, but the process and context through which that ‘thing’ came to be is what really matters.

“The people writing this paper know nothing about what processes led to the genomic transformation they encountered. They also know nothing about the processes (ecological, evolutionary, social) that these things may influence.

“They are content with describing the ‘thing’ instead of the process. At this level, they are right, just as defenders of crop genetic engineering are right when they make up true but irrelevant arguments to avert scrutiny of the safety of their work. They say, ‘DNA is DNA is DNA’, as if the chemical reality of the molecule was all that we needed to know — without acknowledging the role played by many other aspects, such as the DNA sequence.

“By this logic, a play by Shakespeare would be equivalent to an article in the tabloid press, on the grounds that both are made up of letters. But that would be nonsense. It is the sequence of letters, the words, sentences and paragraphs, and the context in which they are all placed that makes each work different from the other and lends a specific identity and function to each.

“This confusion between ‘thing’ and ‘process’ has been there all along since a policy decision was made in the US defining by decree and against reason that GMOs were ‘substantially equivalent’ to non-GM crops on the basis of their chemistry, not the biology of the transgenic manipulation.”