Principles contain useful elements but some serious pitfalls. Report: Claire Robinson

A coalition of US-based NGOs has published an article in Nature Biotechnology putting forward a six-point framework for the "responsible governance" of gene editing in agriculture. The authors state that the principles "build upon lessons learned from the deployment of GMOs", which created public mistrust and consumer avoidance, and instead propose "a more trusted, inclusive and tailored regulatory approach to gene editing".

The authors, from the Center for Science in the Public Interest, Consumer Federation of America, Environmental Defense Fund, National Wildlife Federation, The Nature Conservancy, and the World Wildlife Fund US with help from the Keystone Policy Center, write that while gene editing can bring benefits, "these technologies also raise potential for societal concerns, environmental and health risks, and conflicts with cultural and spiritual values".

They warn that the US has "inadequate regulatory oversight to address concerns presented by biotechnology" and propose that "Regulatory bodies need to assess the potential risks of a particular use or application of gene editing using a science-based risk assessment".

The article, couched as an "invitation" from the NGOs for a wide range of stakeholders to engage in developing regulations for gene editing, contains elements that we at GMWatch support and welcome. But it gets some fundamentals badly wrong, resulting in an article that could mislead regulators and the public about what responsible governance of gene editing would look like – and about the standards that the global NGO community wants to see upheld.

Gene-edited products can't be assumed to be as safe as conventional ones

Among the strengths of the article, the authors issue a caution that flatly contradicts claims from industry and some regulators that gene-edited products are no more risky than conventionally bred ones: "In contrast to the current US Department of Agriculture (USDA) rules, the risk associated with gene-edited products should not be equated to that from conventional breeding."

They also rightly draw attention to the unintended effects of gene editing, including off-target effects in the genome and potential negative environmental impacts.

GMWatch welcomes these elements, which are essential to maintain protections for health and the environment. But we offer some qualifications and criticisms of other elements of the article, which, if not addressed, would seriously undermine any attempt to establish protective regulations.

1. Importance of labelling is missed

The authors make no statement in favour of labelling. They do, however, state that “The public should have access to clear information identifying which gene-editing applications are in use in food, agriculture and the environment”, in order to ensure "basic public information and monitoring capabilities".

Their solution to providing transparency is an online public register of gene-edited products – something that GMWatch and others fully support and have also demanded. We agree that a register is vital, but it is no substitute for clear on-package labelling to inform the consumer as they shop. Both are needed.

2. Authors fail to see that tiered regulation is a trap that favours the industry

Tiered regulation for gene-edited organisms is suggested ("Regulation should be based on the best scientific evidence available and should be tailored and proportionate to the likelihood of risk. Use a tiered approach to align the assessment with the likely risk associated with a gene-edited product”). But tiered regulation in the case of gene editing technology is extremely dangerous, given our total lack of knowledge of the risks, both short- and long-term, of gene-edited plants or animals – let alone microbes, a class of organism that is poorly understood.

The pesticide industry loves tiered regulations and there is a good reason for that. If they are able to argue at Tier 1 (typically in vitro tests on certain types of isolated cells) that they don’t see any adverse effect, then they don't have to carry out further "higher tier" tests. Typically, higher tier tests include experiments in a whole living animal. Without such tests, important adverse effects would be missed and we would in effect be in a "Wild West" position of zero regulation.

By analogy, taking apart the engine of a car and looking at individual parts in isolation does not tell you about the performance of the whole car. The whole is greater than the sum of the parts, and no amount of scrutiny of cell lines or individual organs is going to tell us about effects on a living organism. This principle is well illustrated by an article by Sharon Lerner for The Intercept, on the US EPA's phase-out of animal testing. It should be required reading for all those concerned about risks from pesticides and GMOs alike.

The problem with tiered testing for risky products is that assumptions can be made at the start of the process that the product in question will be safe, based on argumentation (theorising) put forward by industry and some regulators. Therefore, the reasoning goes, there is no need to subject the product to more in-depth tests. But whether we are talking about chemicals or gene-edited GMOs, these assumptions will likely prove wrong. This is especially so with new GMOs, given our total ignorance of their health and environmental effects.

Any tiered regulation would almost certainly be based on assumptions like "small changes in the genome mean small risk". Such an assumption would be wrong, as scientists independent of the GMO industry have repeatedly pointed out that large risks can result from small changes.

The authors' recommendation of “proportionate” regulation for gene editing should ring alarm bells because it is exactly what is being demanded by industry lobbies. When it comes to assessing risky technologies, "tiered" and "proportionate" regulation belong in the waste bin. ALL GMOs, without exception, should be subjected to thorough testing and risk assessment. Tests applied must include generic compositional analyses using cutting-edge "omics" techniques, animal toxicity feeding trials, rigorous allergy screening, and environmental risk trials.

3. Route to proper assessment of risks is missing

The authors rightly say, "risks may be associated with unintended non-target changes in the genome or environmental impacts of toxicity, gene flow and invasion. Where the potential risks associated with a product are substantial and cannot be adequately managed, that product should not be approved for use and released into the environment.”

GMWatch agrees – but we will only stand a chance of assessing these risks adequately if we subject every GMO to a comprehensive risk assessment, as shown above.

4. Authors back industry's demand for product-based risk assessment

Based on the example of the gene-edited hornless cattle which US FDA scientists found to unexpectedly contain antibiotic resistance genes, the authors correctly infer that "the risk associated with gene-edited products should not be equated to that from conventional breeding".

However, they perversely also conclude that "the product, rather than the process used to develop it, should be the subject of evaluation”.

We can only speculate that the authors' lack of familiarity with, or expertise in, the GMO field or certain details of molecular biology leads them to make this mistake in logic. Clearly it was only because the US FDA scientists who discovered the antibiotic resistance genes knew that gene-editing *processes* were used to develop the gene-edited cattle, that they conducted the analyses needed to identify the rogue genes, which were introduced inadvertently during the process.

If the FDA scientists had only focused on the product, as the authors of the new paper recommend and as pro-industry lobbyists have consistently demanded, they would have assumed that the animals were simply hornless cattle just like any naturally bred hornless cattle. They would not have bothered to look more closely at how the animals' genetic sequence had been scrambled by the gene-editing processes. And knowledge of the particular genetic engineering processes used is necessary to inform regulators what to look for, in terms of unexpected effects on the genome.

In truth, what we need is regulation that is both process-based and product-based, but it must start with the process, not with simply looking at the intended product.

5. "The people's GMOs" do not exist and never will

The article falls victim to what GMWatch calls "the people’s GMOs" narrative. The authors demand "robust, inclusive societal engagement" in gene-edited organism development: "Technology developers should take proactive, inclusive steps to identify, engage, and incorporate input of diverse communities, stakeholders, and publics at an early stage in the research and development cycle of individual products." They add, "Product developers should consider, with the input of societal stakeholders, how gene editing could be applied to support a broad range of benefits.”

GMWatch agrees that early public engagement is crucial, but the authors word this in such a way that suggests that the solution to all problems is some form of gene-edited GMO. In order to avoid this "Hobson's choice"*, public engagement must occur from the very beginning of discussions to identify the real problems faced by our food and agriculture systems. Input should be invited from experts from a wide range of disciplines, including ecologists, agroecologists, farmers, consumers, and the non-GMO and organic sectors. Once genuine problems are identified, we need to consider a range of solutions, which could potentially include GMOs, and choose the most desirable options in line with our societal and ecological goals.

For example, for the management of saline soil, do we want a gene-edited GMO saline-tolerant plant (bearing in mind that these do not yet exist in commercialisable form despite years of research investment), or a non-GMO saline tolerant plant, or an agroecological method of improving the soil so it is no longer saline?

6. GMOs will never be inclusive

Still on the "people's GMOs" theme, the authors call for "Inclusive access to gene-editing technology and resources can help drive societal benefit”.

While it's fine to dream, the authors should make themselves aware of research, such as that in a recent Testbiotech report, showing that Corteva controls most patents on CRISPR gene editing, giving it almost monopolistic control over the technology as it's applied in agriculture.

Gene editing, like all GM tech, is a big boys’ game and small players will never perform a significant part. If a small company comes up with a marketable product, the company is almost certain to be taken over by Big Ag or the product will be licensed to Big Ag. This is the way the biotech market has always worked and will continue to work. And because it's a profitable model for all involved, including the small players who get to sell or license their product to the big players, there is no incentive to change it.

Even if this were not the case, the "people's GMOs" fairytale is always going to be bedevilled by the 25-year-long failure of GMO developers to come up with even a single product that has provided "societal benefit".

More organisations should get involved

Despite these issues with the article, GMWatch welcomes the authors' invitation to engage in the process of developing protective regulation for gene-edited products and hopes that a wider range of concerned organisations get involved.


* A supposedly free choice in which only one thing is actually offered.