Earlier this month, the United States Food and Drug Administration (FDA) published its analysis of what has been touted as the “Poster Child” of gene-edited animals: cattle genetically engineered (GE) to prevent them from developing horns. As mentioned in an FDA press announcement released the same day as the article, FDA scientists detected “previously unreported, unintended alterations in [the] genome-edited bulls.” Steven M. Solomon, DVM, MPH, director of the FDA’s Center for Veterinary Medicine, went on to say that while “existence of [the] unintended alteration does not necessarily mean that the genome edit is unsafe to animals or consumers, it does show that both scientists and regulators need to be alert to the potential for such unintended alterations to take place.”
But both scientists and regulators should be well aware by now that unintended alterations can take place when using lab-based biotechnological processes to insert genes into, or otherwise alter existing genes in, living organisms. For decades scientists have known that these biotech processes can cause unintended insertional mutations in recipient organisms, unintended scrambling of recipient DNA at insertion sites, unintended rearrangements of and mutations in the inserted DNA itself, and even (in one case at least, NK603 maize) creation of an unintended combination of inserted and recipient DNA that is transcribed into a chimeric mRNA.
Granted, because the gene-editing processes are newer, we know less about unintended consequences that are specifically associated with these techniques. But even for gene-editing processes we’ve already known for years that unintended “off-target effects” can accompany use of these techniques as well.
The Nature of the Hornless Cow Mistake
This case of the GE hornless cattle illustrates a situation worse than simply not being “alert to the potential” for the gene-editing process to result in unintended consequences, however. The developers of these GE cattle had proclaimed that only the DNA sequences they intended to insert—DNA isolated from cattle—had in fact been inserted into their GE product; they claimed that their GE cattle were therefore “100% cow” and contained no DNA from any other organism or source. And they used this claim to seek FDA’s permission to have humans consume their GE cattle.
But the developers of these GE cattle had never verified this claim of theirs that the genomes of their GE cattle were 100% cow. So in essence, their claim was only a hypothesis; they hypothesized that their gene-editing process had worked exactly as they had hoped it would and that, as a result, only the DNA they’d isolated from cattle and intended to insert into their GE cattle “products” had in fact been inserted.
Is that a big mistake?
Yes, I believe that is a big mistake for any scientist to make. And it’s an especially big mistake if one aims to demonstrate the “precision” of a new technology to a regulatory agency and thereby hopes to influence regulatory policy regarding that new technology.
It’s a big mistake because, as every scientist knows (or should), a hypothesis is only the start of the scientific process of fact-finding. The next step is to gather and analyze data that will prove or disprove the hypothesis.
And the developers of these GE cattle actually had the necessary data in hand that would have disproved their “100% cow” hypothesis. They had the data in hand and yet they did not use it to verify (or not) their hypothesis; instead they went to FDA with only an untested hypothesis, asking the agency for permission to feed their incompletely analyzed experiment to humans.
Special Care Should be Taken in Developing “Poster Child” GE Products
I would have expected the developers of the GE hornless cattle to be especially careful in producing and analyzing what had been expected to be the world’s first commercially available gene-edited animal. And yet they didn’t even use data they had readily available to verify a claim they hoped would convince FDA that there is no need to regulate GE products like theirs.
Similarly, I would have expected the developers of Golden Rice to be especially careful in producing that “poster child” for GE products produced for humanitarian reasons. And yet, the developers of Golden Rice also failed to use DNA sequencing data they had in hand to avoid unintended mistakes attributable to the genetic engineering process itself. They failed to screen their potential GE products for those “events” in which insertional mutations had been created by the genetic engineering process; as a result of this mistake development of Golden Rice was delayed by at least half a decade.
Implications for Other GMO Products
Such big mistakes made during the development of these “poster children” GMOs make one wonder how carefully other, more run-of-the-mill GMOs have been developed over the last 25 years as well.
They also justify—in my mind at least—FDA’s intention to “carry out mandatory premarket review of all gene-edited livestock” and thereby regulate these GE products based on the fact that they were created using genetic engineering. (I also think FDA should require premarket review of all GE crops as well.) We know the genetic engineering process itself can result in certain types of unintended consequences; we also know that (at least some) developers of GMOs are not being sufficiently careful about analyzing their putative GE products to identify the known types of unintended alterations; therefore, FDA’s decision to require premarket review of all gene-edited livestock to ensure that developers are conducting careful, thorough analyses of their potential GE products prior to commercializing them is a logical and science-based conclusion for the agency to have reached.
Nevertheless, the editorial staff at Nature Biotechnology published their own piece in the same February 7, 2020 issue of the journal as FDA’s analysis of the GE hornless cattle which encourages FDA to regulate gene-edited livestock based only on risks associated with the trait introduced into the GE animal (e.g. “hornlessness”) as opposed to regulation based on the process through which those traits were introduced.
The Nature Biotechnology editorial staff related some of the same old arguments I’ve heard for 25 years, like: the public doesn’t have a problem supporting GE drugs so why shouldn’t they support gene-edited animals as well? My response has been the same for 25 years: sick people are willing to take more risks to regain their health than healthy people are willing to take with the food they feed themselves and their families.
The editorial also brought up a new, but disingenuous, argument: “In over 30 years, there has yet to be a single compelling case where genetic engineering posed a safety problem in a food animal.” It’s disingenuous because…GE food animals have not been around for 30 years; in fact, as stated later in the Nature Biotechnology editorial, the “world’s first genetically engineered animal…still has not reached US consumers.”
The Bottom Line For Me Is:
Scientists still have a lot to learn about unintended consequences associated with all methods currently used to genetically engineer organisms, especially newer methods like gene-editing;
(At least some) developers of GMOs aren’t being careful enough in checking their GE products for the unintended consequences of GE processes/methods that we already know about—even when those products serve as “poster children” for gene-edited animals or humanitarian uses for GE crops;
Therefore, all organisms produced using these GE processes/methods should be regulated by federal agencies on a case-by-case basis to ensure that developers are being scientifically rigorous in their analyses of their products and that possible risks to the US food supply are identified prior to product commercialization.
And the FDA is on the correct course…at least for gene-edited animals…in spite of the Course correction published by the editorial staff at Nature Biotechnology.