ashley p. taylor is a Brooklyn-based journalist specializing in science and technology.
Illustrations by Peter Bollinger
Published July 29, 2019
In February, a restaurant-chain in the Midwest started using oil derived from gene-edited soybeans for frying and in salad dressings and sauces. Calyxt, a Minnesota-based company, edited the genome of the soybeans to give the oil higher levels of oleic acid, which, the company says, boosts levels of good cholesterol. According to the company, processed foods containing the oil (brand name: Calyno) should hit supermarket shelves soon. Calyxt’s high-oleic soybean oil is the first gene-edited food product (as opposed to “genetically modified organism”; more on that later) to be sold commercially in the United States, having received approvals from both the Department of Agriculture and the FDA.
For years, technologists have talked about how precise, low-cost approaches to gene editing, such as TALENs and Crispr-Cas9, could change the world. And there are good reasons to believe that the world-changing is about to begin. Many more gene-edited crops are in the works — everything from high-fiber wheat to bruise-resistant potatoes to a domesticated version of the vitamin-packed ground cherry, a relative of the tomato that tastes more like a berry but whose unruly growth has so far prevented commercialization.
The goals of those engaged in gene editing of crops vary. Some researchers are attempting to improve food security in a world with an ever-growing population. Others want to create foods that are cheaper or healthier for commercial gain.
The primary reason those in agricultural biotechnology (agribiotech) generally choose gene editing is straightforward: it is faster, easier and more accurate than other methods of altering genomes. But there are other factors. To date, Washington has not applied some of the costly and time-consuming regulations that have constrained development of genetically modified organisms (GMOs) since the 1990s. Agribiotechs are also banking on the idea that consumers (as well as governments) will see gene-edited plants as being fundamentally different from GMOs — and will avoid a repeat of the GMO consumer backlash.
A Brief History of Genetic Engineering in Crops
Long before humans had a clue about the mechanisms of genetics, they were influencing the genomes of foods by fostering crops with desirable characteristics of taste, size, drought tolerance, orderly growth and so on, while abandoning others and crossing strains with varying positive traits to generate offspring that contain them both. This applies even to organic crops today, in the sense that all have had their genomes modified by humans through selective breeding. Think seedless grapes and sweet corn.
These first GMOs had traits that benefited farmers. Little thought was given to consumers and how they might react to the idea of eating plants whose genomes contained DNA from other species.
In the 1970s, scientists began to figure out how to modify genomes in the lab, cutting and pasting stretches of DNA from different species and assembling new sequences in what’s called recombinant DNA technology. A common approach was to take a virus or bacterium that modifies host genomes during infection and reprogram it as a vector to insert the scientist’s gene of choice, called a transgene (organisms modified this way are called transgenic). For plants, the most commonly used vector is Agrobacterium tumefaciens, or crown gall bacterium, which infects a variety of plants — giving them “galls,” or warts — and in so doing inserts some of its DNA into their genomes.
Beginning in the 1990s, Monsanto (acquired by Bayer in 2018) and others unveiled varieties of transgenic corn and soybeans that produced their own insecticides or were resistant to the widely used herbicide glyphosate, which Monsanto sold as Roundup. In the latter case, this allowed farmers to kill weeds in their fields without harming their “Roundup Ready” crops.
These first GMOs were designed to meet farmers’ needs. Little thought was given to how consumers would react to the idea of eating plants whose genomes contained DNA from other species, explains biotech expert Vonnie Estes, who has worked for both Monsanto and Caribou Biosciences. “At the time, people just didn’t think anyone would care,” says Estes. “It was the arrogance and naïveté of thinking, ‘Oh, we did this great thing and everyone should just accept this.’”
But many consumers did care. There was a backlash to GMOs in the form of boycotts, vandalism and political initiatives to require labeling. So while transgenic GMOs are now thriving — over 90 percent of corn and soy grown in the U.S. is GMO — the backlash has probably prevented development of other GMO plants. “GMOs could have been used in all different kinds of crops, but companies just didn’t spend the resources doing it because they didn’t think consumers would accept them,” says Estes.
Whatever the validity of consumers’ fears, from a scientific perspective, one problem with generating transgenic crops is that it is impossible to control where in the genome the transgene lands. The great advantage of gene editing is that it allows scientists to make changes at sequence-specific locations rather than at a random spot (see page 36). It’s much more precise. This is key. Of the several methods for gene editing, Crispr-Cas9 is the easiest and the one getting the most buzz, but there are others. Calyxt, for example, uses a gene-editing approach called TALENs.
Agribiotech pioneers are hoping that, in addition to these scientific advances, they can start off on the right foot with the public so that gene-edited crops don’t meet the fate of the first GMOs. “Those lessons were hard learned,” says Estes.
Companies are also avoiding using terms like “GMO” and “genetic engineering” in reference to their gene-edited crops. Calyxt is calling the new technology “precision plant breeding” to generate “non-GMO” crops, emphasizing that its products are not transgenic and do not contain “foreign DNA.” These terms, which seem technical on the surface, send a political message: the new breeds are not the GMOs you once protested. And, indeed, they are not. But they are genetically engineered plants, and politics is bound to reflect that reality.
Feeding 10 Billion
Whatever you call these agribiotech advances, there’s no getting around the need for technology that will make it possible to feed the global population that is expected to approach 10 billion by 2050 (compared to 7.6 billion today). What’s more, rising incomes virtually guarantee that consumers will demand disproportionately more animal protein, which uses more resources than plant foods to yield equivalent nutrition. All told, that means food production will probably need to double.
Food security is also threatened by climate change, which means more droughts, floods, heat waves and violent winds. Farmers and food supplies will also face increased plant pests and diseases, since these will spread to new localities as the planet warms and rainfall patterns change.
Climate change could also increase threats from plant pests and diseases, as these will advance into and thrive in new locations as the planet warms and rainfall patterns change.
In fact, this state of affairs is no longer hypothetical. The carb-packed root vegetable cassava is essential for food security in much of sub-Saharan Africa. But the crop is threatened by cassava brown-streak disease, which renders the root inedible and is transmitted by the whitefly. And as the climate warms, the whitefly’s range is expected to expand. Cassava brown-streak disease is already epidemic in eastern and central Africa, but has yet to spread beyond.
The disease hasn’t reached Nigeria, which produces more cassava than any other country in the world. But “when it gets there, it’s going to be a big problem,” says Nigel Taylor, a plant biotechnologist at the Donald Danforth Plant Science Center in St. Louis. Taylor and colleagues are collaborating with researchers at the University of California-Berkeley to use Crispr-Cas9 to create plants that are resistant to cassava brown-streak. They have introduced mutations into two genes in the plant that encode proteins the viruses must hijack to cause disease. So far, the gene-edited cassava plants are partially but not entirely resistant to brown-streak.
Taylor is also planning to use gene editing to increase the productivity of teff, the most important cereal crop in Ethiopia. During the Green Revolution of the 1960s and ’70s, agronomists used selective breeding to develop shorter-stemmed varieties of wheat and rice that were sturdier and less likely to fall over and spoil. Since these varieties put less energy into growing tall, they also put more energy into producing grain and thus have higher yields. Taylor wants to use gene editing to create short-stemmed teff. “I think we can do it, and do it quickly,” he says.
Taylor’s goals are entirely humanitarian; he’s funded by organizations that include the Bill & Melinda Gates Foundation. The products he is developing would be free to the farmers who need them.
But for-profits are also tinkering with plant genomes in an effort to do well (financially) while doing good. The Massachusetts-based company Yield10 Bioscience, for example, aims to make crops more productive. Yield10 is developing a variety of the Camelina sativa (a versatile oilseed notable for its high content of omega-3 fatty acids) in which three genes have been inactivated through Crispr in order to increase the oil content of the seeds. They’re also working on a canola plant that is gene edited for higher oil content, and on rice and wheat that are edited for higher yield.
One note of caution shared by many of the experts: gene editing alone is unlikely to resolve the world’s food security problems. “It’s going to take a systems-based approach,” says Yield10 CEO Oliver Peoples. “My view is it’s really going to be about selecting the right seed with enhanced-yield genetics developed through genome editing or GMO with the right seed-coating package, and planting it with the right combination of fertilizers and nutrients at the beginning of the season,” he says. But there’s only so much one can control. “As the season progresses, you’re pretty much at the mercy of the weather,” Peoples adds.
Beyond efforts to ensure there is enough food for future generations, agribiotech companies are using gene editing to tailor edibles to special needs — think vitamin-packed foods and non-allergenic versions of foods. Not to mention improved aesthetics, as in fruits and veggies that don’t turn brown when cut open. This interest in niche markets wasn’t commercially practical with the first GMOs because regulatory costs were prohibitive.
Rather than forgetting to consider consumers — the mistake Estes says companies made with the first gmos — calyxt seems to be imagining their consumers down to the grocery stores they frequent!
Initially, only big companies aiming to tap mega-markets could afford to do genetic engineering. That’s why the original GMO plants were row crops, notably corn and soy, that are used in animal feed and in processed food. Since the vast majority of these crops aren’t sold directly to consumers (though some sweet corn is GMO), companies didn’t engineer them with consumers’ sensibilities in mind.
By contrast, the new gene-edited crops have so far been subject to less regulation than the original GMOs — at least in the United States — and are generally less costly to create. One result: startups can afford to get into the business of genetic engineering and work on crops that do not have the potential to make mega-bucks.
“Now you can tap into that market where you’re going after niche applications — a modified oil or modified starch or a non-allergenic peanut,” for example, says Peoples. But the focus is not on farmers. “The value proposition is being able to offer the consumer a product that’s either healthier or safer.”
Take Calyxt’s gene-edited oil. Most soybean oil is hydrogenated to make it more shelf-stable, a process that generates trans fats. Yet there is now a medical consensus that trans fat consumption increases the risk of heart disease. Calyxt’s high-oleic oil deteriorates more slowly than its non-edited counterparts, according to the company, which means that it does not require hydrogenation and is trans-fat-free. The oil meets FDA guidelines for a “heart-healthy” claim, meaning that products containing the oil may, depending on their other ingredients, be eligible to have a heart logo on their packaging and thus appeal to health-conscious consumers. Such a logo may go “a long way in convincing a consumer at the grocery store,” says Calyxt’s chief commercial officer Manoj Sahoo. Rather than ignoring the views of consumers the way the GMO makers did, Calyxt seems to be keeping its eyes on the prize.
Actually, there may be a lot of prizes out there. For example, scientists are trying to use gene editing to create a hypoallergenic peanut — a goal that, if realized, could allow peanuts back on the menus of schools, airlines and sports stadiums. Researchers are also working on gene editing to generate wheat that would be safe for people with celiac disease.
On the niche end of the spectrum, consider the aforementioned ground cherry, which biologist Joyce Van Eck, of Ithaca’s Boyce Thompson Institute, and Zachary Lippman, of Cold Spring Harbor Laboratory on Long Island, are domesticating using gene editing. It’s unlikely the ground cherry would ever compete with, say, apples and grapes for supermarket shelf space. But it is a testament to the efficiency and simplicity of gene editing that researchers are working on a new competitor for the likes of huckleberries.
The North Carolina-based agribiotech company Pairwise, for its part, hopes to take gene editing in a different direction, making foods that are more convenient to eat — though the company isn’t specifying exactly what kinds of products they’ll come out with. (I, for one, vote for the seedless lemon and an easy-to-peel navel orange.)
Regulation seems to sneak its way into every aspect of analysis of gene-editing technology. And for good reason: as already noted, regulation affects not just how crop developers do their work, but where they focus their efforts, based in turn on what they can afford.
With genetically engineered plants, federal regulation means much more than oversight. When the USDA’s Animal and Plant Health Inspection Service (USDA-APHIS) deems a genetically engineered plant to be a “regulated article,” the crop can only be grown in small field trials. To petition USDA-APHIS to get a crop deregulated, the innovator must run a gantlet of costly testing.
It takes an average of 13 years and $130 million or more to bring a new GMO to market, $30 million of which are the costs of complying with regulations — or so says a 2011 survey commissioned by CropLife International, a plant-science trade association. “That puts it in the same ballpark as looking at getting a new drug approved,” says Donald MacKenzie, executive director of the Danforth Plant Science Center’s Institute for International Crop Improvement.
So when it was first announced in 2016 that USDA-APHIS would not regulate a non-browning mushroom whose genome had been edited using Crispr, it was a big deal. Industry insiders began to speculate that gene-edited crops, in general, would not be regulated the way so many transgenic GMOs had been.
Escaping regulation is indeed a big deal for agribiotech, but gene-edited plants won’t do so simply because they are gene edited rather than transgenic. There are transgenic plants that have received non-regulated status from USDA-APHIS, and there are some cases, though they remain theoretical, in which USDA-APHIS would likely regulate gene-
edited plants. In fact, the same policies, be they from USDA, FDA or EPA, cover all genetically engineered plants. It just so happens that because of the way gene editing is done, it hasn’t (so far) triggered the same USDA-APHIS regulations that some — but not all — transgenic GMOs do.
In short, USDA-APHIS regulates any genetically engineered plant that contains genomic sequences from a plant pest, was created using a plant pest as a vector or has potential to be a plant pest itself. Those seeking to cultivate genetically engineered plants for general use must submit what are called “Am I Regulated?” letters to USDA-APHIS describing each product and making the case for its non-regulated status. Both the letters and APHIS responses are publicly accessible online and provide a sort of case history of USDA-APHIS’s regulatory practices.
Early GMOs often used Agrobacterium tumefaciens as a vector for transgene insertion and, because of that, had some Agrobacterium DNA permanently incorporated into their genomes. These products automatically triggered regulation because the vector is a plant pest. In addition, in some cases, DNA from Agrobacterium or from other plant pests, such as cauliflower mosaic virus, is included to help direct expression. If it remains in the genome of the final product, the crop is by definition a “regulated article,” regardless of the vector used to insert the transgene. Finally, if the transgene itself comes from a plant pest, as is the case for the gene that confers glyphosate resistance in Roundup Ready soy, regulation is automatic.
Other transformation methods, such as the gene gun, have allowed scientists to insert DNA into genomes without using Agrobacterium as a vector, and transgenic plants generated this way aren’t necessarily regulated. For example, in 2014, the now-defunct San Francisco-based Taxa Biotechnologies received USDA-APHIS approval to produce an ornamental plant “genetically engineered to emit a pleasant, dim glow.” The company’s efforts to create a glowing plant failed for financial reasons, but it planned to insert synthetic versions of genes from fireflies and jellyfish via the gene gun.
Though not a food, the glowing plant is just the sort of thing to rouse anti-GMO activists. Yet USDA-APHIS decided not to regulate it because it contains no plant pest sequences. Scotts Miracle-Gro’s Kentucky bluegrass, which was transformed with a gene for glyphosate tolerance using the gene gun, is another transgenic crop that is not regulated by USDA-APHIS because it contains no plant pest sequences.
USDA-APHIS regulates any genetically engineered plant that contains genomic sequences from a plant pest, was created using a plant pest as a vector, or has potential to be a plant pest itself.
Further, in some cases scientists use Agrobacterium as a vector for genetic engineering but leave no DNA from Agrobacterium or other plant pests in the genome of the final product. This is the case for many gene-edited crops: Agrobacterium is used to insert the gene-editing machinery into the plant. But once the edits are made, the editing machinery, including any related pest sequence, is removed. That leaves only the edit — the kind of genetic change that could have happened by chance or through conventional breeding.
Intricacies of Editing
Not to beat a dead horse, but USDA-APHIS has been clear from early on what sort of gene editing would trigger regulation. Back in 2011, Cellectis, Calyxt’s parent company, wrote to USDA-APHIS describing a non-Crispr gene-editing method and asking if crops modified in this way would be regulated. USDA-APHIS replied that in cases where the gene editing results in small insertions and deletions without introducing any DNA, the products would not “in most cases” be regulated. Cases of gene editing where scientists introduced DNA, either as a template for editing or as a transgene, would be considered on a case-by-case basis.
So far, one crop whose genome was edited using introduced DNA, a disease-resistant corn variety, has been through the “Am I Regulated?” process and has received non-regulated status. As of this writing, more than 20 gene-edited crops have been granted non-regulated status. But these crops are not avoiding regulation because they are gene edited.
The EPA regulates pesticides, so any crop that produces its own bug dope, such as Monsanto’s controversial Bt corn, which has been genetically engineered to express an insecticide gene from the bacterium Bacillus thuringiensis, is subject to regulation. The FDA, for its part, holds voluntary consultations with biotech companies about new products created through genetic engineering. Calyxt’s high-oleic soybean oil was the first gene-edited crop to undergo such a review; in the end, the FDA provided its stamp of approval.
In Europe, where regulations on genetically engineered crops are more stringent than in the United States, the European Court of Justice has ruled gene-edited plants are to be regulated the same way as transgenic GMOs — a disappointment to business. Many scientists share the view: the European Commission’s own science advisors issued a statement in 2018 essentially opposing the court’s decision, suggesting it would be better to focus on the safety of agribiotech products rather than on the processes by which they were made.
Says Danforth’s MacKenzie, the ruling “has, I’m sure, had quite a chilling effect on the prospects for innovation and technology development in Europe.” The court decision, by the way, also applies to companies outside the E.U. contemplating sales to Europe.
Regulation also raises the issue of collateral damage. Even if Nigel Taylor’s gene-edited cassava was completely disease-resistant and ready to start field tests in African countries that grow cassava and have cassava brown-streak disease, the innovators will be out of luck if their African hosts choose to model their own regulation of gene editing on that of Europe.
Representatives of agribiotech companies will tell you, again and again, that gene editing makes no changes that could not be achieved through traditional breeding or that could not occur in nature. The goal is to distinguish their products from transgenic GMOs like Bt corn. But this isn’t entirely true. No, the insertion of a gene from a bacterium into a plant would not occur as the deliberate outcome of traditional plant breeding. But genes, including those from bacteria, do jump from one species to another without human intervention in a process called horizontal gene transfer. With that, plus mutation, there’s a very small probability of any novel, and useful, genomic sequence turning up.
Lawmakers use these terms and phrases, too, and do so in different ways. For example, in the U.S., the National Bioengineered Food Disclosure Law (2016) requires manufacturers to label products containing “bioengineered” foods — defined as those containing genetic material modified by recombinant DNA techniques and whose modifications “could not otherwise be obtained through conventional breeding or not found in nature.”
The final ruling on how the law is to be implemented does not explicitly state that gene-edited foods will be subject to the disclosure requirements. In other words, gene-edited plants are exempt from the labeling requirement on the rationale that their mutations are “natural.” The European Court of Justice, on the other hand, decided that the changes produced by gene editing could not have occurred in nature and that, therefore, gene-edited crops should be regulated. These opposite interpretations of what could occur in nature just go to show that regulations distinguishing between gene-edited and transgenic crops are mired in semantics.
The use of gene editing to make crops more efficient to grow could reduce the cost of generating the product, but whether those savings will go to companies, farmers or consumers remains to be seen.
A similar confusion arises around the term “GMO.” Agribiotech companies doing gene editing insist their crops are “non-GMO.” But common sense would tell you that to edit a genome is to modify it, and therefore gene-edited plants are GMOs. To accept that, however, might put regulators on a slippery slope. Nature also modifies organisms through mutation, suggesting that just about every organism is a product of genetic modification.
Since the distinctions between gene-edited crops and other GMOs and “bioengineered” foods are weak, it seems likely that, as a general principle, whatever rules apply to transgenic GMOs should also apply to crops that are modified by gene editing. That may rid the analysis of semantic gymnastics. But it would also close off the path to swift development and commercialization of foods created using agribiotech.
The compromise would be to regulate genetically modified crops, be they transgenic or gene-edited, but regulate them less — and then label them as such. That way, people who don’t care how a crop was created could benefit from the changes, while those opposed to genetic modification could go without.
Who Gets the Gravy?
Will consumers see the benefits of gene-edited foods in prices at the supermarket? Compared with agribiotech products that are regulated by USDA-APHIS or the EPA, gene-edited crops are bound to be less expensive to get to market. Moreover, gene editing that makes farming more efficient by raising yields or guarding crops against environmental stresses could reduce production costs. But just how the savings will be divided among farmers, consumers and the owners of the technology remains to be seen, says Estes. If gene-edited foods have traits that consumers particularly value, companies will probably charge more for them, she notes.
Currently, most GMOs are used for animal feed or in processed food. (Disease-resistant papaya from Hawaii and a non-browning apple — both transgenic GMOs — are exceptions.) If foods like the ground cherry are domesticated and commercialized with gene editing, perhaps consumers will start to see GMOs as normal rather than as frankenfoods.
The marketing focus on how gene editing will increase variety as well as make foods cheaper and healthier is no accident. “If we can really lead with the benefits to consumers,” Estes opines, “then I think there is a category of consumers who are going to be more interested in the benefit and less interested in how it’s derived.” The goal, of course, is to avoid the GMO debacle.
How it Works
The first genetically modified organisms were made using transgenesis, in which a gene from one species, called a transgene, is inserted into the genome of another. The most common method for getting transgenes into plant DNA is to use the crown gall bacterium Agrobacterium tumefaciens. In the wild, Agrobacterium inserts some of its own DNA into plant genomes when it infects them. For genetic engineering purposes, scientists modified the Agrobacterium genome to insert not its own DNA but a transgene of choice in a process called transformation.
Transgenes can also be inserted using biolistics, in which scientists use what’s called a gene gun to shoot DNA-covered gold pellets at plant cells. Alternatively, the transgenic DNA can be inserted using a technique called protoplast transformation.
It is not possible to control where in the genome the transgene ends up with any of these techniques. Gene editing, on the other hand, allows sequence-specific genomic changes. With gene editing, a DNA-cutting protein binds to DNA in a sequence-specific way and makes a cut. Afterward, cells use their own methods to repair the breaks, which can result in changes, or edits.
Sometimes the cellular machinery just sticks the DNA ends back together, a process that often results in the random insertion or deletion of DNA letters, which can completely change the protein that the gene encodes, often rendering it non-functional. If the goal of the editing is to turn off a gene, this is a good way to do it.
The break can also be repaired using DNA that scientists have introduced, which may differ from the original DNA in some way, as a template. This method – similar to cutting and pasting – is a way to edit a gene without turning it off. In theory, it could also be used to insert a transgene.
There are several methods for editing genes. Two earlier methods, TALENs and zinc finger nucleases, use proteins that bind to particular DNA sequences – just a few base pairs long – for DNA recognition. To generate a molecule that recognizes a longer sequence, scientists string together several of these proteins. Then, to make the system cut the DNA, scientists fuse the DNA-recognition proteins to a cutting protein. But generating these very specific proteins to recognize and cut a DNA sequence is not easy, and these methods required engineering a new protein for every sequence to be edited.
The relatively new gene-editing method with the acronym Crispr does not use a specially engineered protein. Rather it employs an RNA molecule called a guide RNA to recognize the DNA sequence to be edited. In Crispr, the guide RNA leads a cutting protein, usually a protein called Cas9, to the location to be cut.
One aspect blurs the regulatory line between gene editing and transgenesis: gene editing with Crispr requires introduction of the cutting nuclease and the rest of the Crispr machinery into plant cells. That process involves inserting Crispr-Cas9 components into the plant genome by Agrobacterium, gene gun or protoplast transformation methods. Scientists then select for resultant plants that have only the intended edits and not the CRISPR-Cas9 transgene.