A Cow, a Controversy, and a Dashed Dream of More Humane Farms

On the morning of August 7, Alison Van Eenennaam awoke to a tweet from a man she had never met. He had sent her a link to a story written in German, illustrated with a clip-art cow next to an udder-pink biohazard symbol. “Aren’t you involved in the hornless cows criticized here by a German NGO?” the man tweeted at Van Eenenaam from nine time zones away. “Can you give us some details on what @US_FDA found?”

Van Eenennaam could not. But not because she didn’t have the details.

For nearly two years the animal geneticist and her team at UC Davis had been meticulously poking, prodding, weighing, and measuring a herd of six young, genetically dehorned Herefords. Born on campus in September of 2017, the calves were gene-editing royalty. Their father, Buri, had been created in a lab in Minnesota a few years before, his genome tweaked by the agtech startup Recombinetics to prevent him from growing horns. Horns are considered a menace in the commercial dairy business and typically get burned off, so the startup had set out to use engineering to make a more humane livestock industry.

A world-first, he and his half-brother Spotigy were an overnight media sensation. “We know exactly where the gene should go, and we put it in its exact location,” Recombinetics executives told Bloomberg in 2017. That year, Van Eenennaam secured a half-million-dollar grant from the US Department of Agriculture to see if Buri’s descendants would inherit his genetic alteration as intended, and to study those animals’ health and dairy-producing potential. (Spotigy was sacrificed in 2016 to analyze his meat for quality. He did not sire any calves.)

So as she stared at Twitter on that recent August morning, Van Eenennaam knew more than anyone about Buri and his extended family. Including the fact that six months before, scientists at the US Food and Drug Administration had stumbled upon a surprise in Buri’s DNA. There had been an accident during Recombinetics’ editing process. And the mistake had been passed down to four of Buri’s six calves, along with the hornless gene. Yet those calves were every bit as healthy as their naturally horned Hereford counterparts, as far as she could tell.

But she couldn’t talk about any of that because she had submitted a paper on it to a journal for peer review. If she discussed it now, her paper could get rejected. So she seethed to herself as she read a line in the German story: “No research has been carried out on the possible consequences for animal health, or whether these additional genes are biologically active.”

Research takes time, especially when it involves animals with a nine-month gestation period. By the time those results get published, the study is in a sense old news. It looks back at the long story of a scientific discovery, like peering at ancient starlight through the barrel of a telescope. In the age of the internet, scientists have grown increasingly impatient with the plodding pace of traditional publishing and sought to shake it up. For the most part, these growing pains have remained cordoned off in conference proceedings, invisible to the general public. But sometimes, clashes in publishing culture can spill over into the real world in unexpected ways, shaping public opinion with wide-reaching effects. In the case of the hornless cows—the would-be poster calves of the future of food—it would turn their story upside down.

Van Eenennaam moved from Melbourne, Australia to Davis, California right as genetic engineering was transforming the agricultural college town. In nearby fields the Flavr Savr tomato was making its way out of the ground and into local grocery stores, to become the first genetically modified food available for purchase.

She and other young faculty decided to harness these new genetic engineering tools to do for livestock what other researchers and companies were doing for crops—giving animals and the foods they produced new traits. They embarked on experiments in cattle and goats to make their milk more nutritious. But then their work hit a snag. The FDA decided that introducing foreign DNA into animals qualified them as veterinary drugs, creating a lengthy and costly approval process that dissuaded many scientists. Funding for transgenic livestock research foundered. And people like Van Eenennaam scraped by on the few remaining grants.

LEARN MORE

The WIRED Guide to Crispr

Then came the gene-editing revolution. Tools like TALENS and Crispr, which allow scientists to alter animals’ genomes without adding foreign DNA, reignited livestock researchers’ imaginations and spurred a new crop of companies, including Recombinetics. Its engineers wanted to introduce DNA from hornless, non-dairy cattle breeds into their milk-and-cheese producing cousins. The company sought out Van Eenennaam to help house and study its creations. In 2015, Buri and Spotigy moved to UC Davis.

A year and a half later, Buri’s sperm was used to artificially inseminate ten horned Hereford dams in the Davis herd. When Van Eenennaam got word that six pregnancies had been confirmed, she sent in an application to the USDA Biotechnology Risk Assessment Grant program. Its mission is to fund projects that help federal agencies evaluate emerging technologies. Van Eenennaam was cautiously optimistic that by collecting reams of data on the calves, her team might help make a case for gene-edited animals to be regulated differently than transgenic ones.

Those hopes didn’t last the month.

On January 19, 2017 the FDA issued a set of draft guidelines, lumping gene editing together with older GMO technologies. The USDA, which oversees gene editing in plants, has decided, in most cases, to not regulate the technology, treating it as a sped-up version of traditional breeding methods. But the FDA reached a different conclusion, that the editing process presents unique risks. What if Crispr or TALENS make changes they’re not supposed to make? What if those mistakes result in unexpected mutations? And what if those genetic changes spread from livestock to their wild relations?

“We were right in the middle of doing our experiments and it caught us off guard,” says Van Eenennaam, who has not been shy about criticizing the FDA’s decision. “We had predicated everything on assuming those animals could go into the food supply.”

If the calves—five males and one female—were considered GMOs, they couldn’t be butchered and sold off through the University’s Meat Lab. For Van Eenennaam, that was key to making the economics of her research work out. Instead, each 2,000 pound animal would have to be incinerated, at a cost of 60 cents per pound. So Van Eenennaam petitioned the FDA for an exemption.

In December 2018, she sent the agency a detailed dossier on the calves—disclosing the results of their physical exams, blood tests, and DNA sequencing. Based on her team’s analyses, nothing looked amiss. The animals appeared healthy, their genomes precisely edited, their furry foreheads completely horn-free. While Van Eenennaam waited for the agency to make its decision, she and her colleagues wrote up their results, and in February submitted them to Nature Biotechnology.

While the UC Davis team typed up their manuscript, scientists at the FDA were combing through the data too. One of them was Alexis Norris, a biostatistician in a division of the Center for Veterinary Medicine. When she joined the agency in July 2018, she hadn’t had much experience with the genomes of cows and pigs and other barnyard species. But she’d gotten really good at analyzing massive amounts of sequence data as part of her graduate and post-graduate work at Johns Hopkins, on the genetic underpinnings of human diseases. At the FDA, she was putting those skills to use screening livestock DNA data for unintended edits.

Part of her work involved developing software specific to that task. Norris’s team had long been planning to test-drive it with the genomes of Buri and Spotigy, because theirs was the largest publicly available dataset for a gene-edited animal. “We wanted to make sure it could handle that size data,” says Norris. “And that our computer wouldn’t crash while analyzing it.”

Earlier this year, they used their software to compare Buri and Spotigy’s DNA to a reference bovine genome. Then, as an extra precaution, they also checked the cow’s DNA against a short sequence of circular bacterial DNA called a plasmid. Recombinetics had used this particular plasmid to ferry the genetic instructions for hornlessness into the cells that would later become Buri and Spotigy. It isn’t supposed to stick around. But neither Recombinetics or Van Eenennaam had ever checked to ensure that was true.

As the FDA’s software began to spit out results, Norris saw the hornless gene, right where it should be. But then, next to it, she saw sequencing data that matched the plasmid. “This was a very unexpected finding,” says Norris. What was supposed to be a quick validation had turned up a major surprise. During the editing process, Buri and Spotigy had accidentally acquired a bit of bacterial DNA. It wasn’t much, about 4,000 letters. But the genetic insertion was enough to make the bulls GMOs by any definition.

It took a few weeks for Norris and her team to feel confident they’d made a real discovery and not stumbled on, say, a flaw in their code. On March 6, they broke the news to Van Eenennaam and her collaborators. According to Van Eenennaam, the FDA revealed the plasmid hadn’t only slipped into Buri and Spotigy’s DNA. It had also been passed down to four of Buri’s six calves. (The FDA declined to comment on the calves, citing confidentiality agreements.)

For Recombinetics, the fallout was swift. The company had crafted a deal with Brazil to start a hornless herd there after government officials determined the company’s cows didn’t require special oversight. The company was readying shipments of Buri’s sperm for export when the FDA alerted its Brazilian counterparts about the bacterial DNA glitch. The Brazilian agency soon re-classified Buri (and any of his progeny) as a GMO product. Recombinetics abandoned the project, as WIRED exclusively reported in August, but the company has plans to revisit it in the future with a plasmid-free cell line.

Meanwhile, Van Eenennaam’s team notified Nature Biotechnology of the discovery. Then they went over the calves’ genomes themselves to confirm the presence of bacterial DNA, and sent their results to the journal. During this time, Norris and her FDA colleagues were also busy writing up their findings. The plasmid discovery had been theirs, after all, and they didn’t want other people to make the same mistake. They included in it another potentially incendiary detail: The cows’ bacterial DNA contained a few unwelcome genes for antibiotic resistance. In July, they uploaded a summary of their work to the preprint server bioRxiv, a somewhat controversial repository for manuscripts that have not yet been peer-reviewed. While hailed by many as a crucial tool for accelerating scientific progress, critics worry that preprints, which are meant to be consumed by fellow scientists, might be misinterpreted by the general public.

The study went largely unnoticed until it was picked up in early August by a German organization called Testbiotech. That’s how Van Eenennaam first learned of the FDA publication. Other stories quickly followed on other anti-GMO platforms. Soon her email inbox and Twitter timeline filled with questions and media requests to comment on the preprint. She grudgingly swatted them all away. Because Nature Biotechnology was still reviewing her manuscript, she couldn’t tell people the small success stories embedded in all this mess: that two of the young bulls had successfully inherited the hornless gene with no bacterial DNA; that they were healthy and hornless; and that with proper screening, making more like them would be easy. But as the weeks passed and the anti-gene-editing rhetoric ratcheted up on social media, she despaired that public sentiment was turning against the cows.

“The timing of this whole thing was just really bad,” she says. But Van Eenennaam, who’s never put any of her own work on bioRxiv, also faults the FDA for speeding ahead with publicizing non-peer-reviewed results on a hot-button social issue. “To put that out there without the context of the rest of the story opened it up to get politicized in the way it did.” The agency, meanwhile, supports open access to federally funded research and tends to leave those decisions to individual researchers.

In one early news report, The New Food Economy detailed the preprint’s findings with a headline that read “FDA Finds a Surprise in Gene-Edited Cattle: Antibiotic-Resistant, Non-Bovine DNA.” That story was later updated to clarify the plasmid was unlikely to be dangerous to the cows or a person who would eat them, because it lacked the promoter DNA necessary to turn on the bacterial genes. It’s fairly common for DNA from germs to hop into other species—the human genome contains its fair share of bacterial and viral genes.

Fyodor Urnov, a leading genome-editing researcher at UC Berkeley, applauds the FDA’s work and says that putting it on bioRxiv was the right thing. But in retrospect, he wishes the agency had coordinated with Van Eenennaam to release their results together. “Then one could see that it is both true that an honest mistake was made, and that we actually know it has no consequence,” he says. Now it’s a matter of trying to straighten out the story. In genetic engineering, he says, contentious information can take on a life of its own in the public eye, “because of the expectations that something is being hidden. I’m sure for some people this looks like exhibit 1A for that.”

Laura Epstein, a senior policy analyst at the FDA, says the decision to publicize the discovery was in part to aid other researchers or companies looking to develop gene-edited animals. “We want them to be able to take this into account,” says Epstein. “We actually think it will help move the field forward and help people use this technology in a way that gets products to market.”

Those products, at least for the time being, won’t include hornless dairy cows. Recombinetics is not seeking approval for the animals in the US. And in May, the FDA doled out its final decision on Van Eenennaam’s calves. None of them could go into the food supply, including the two bacteria-free males. As soon as the UC Davis veterinary school has an opening for them, they’ll be incinerated, joining their three GMO brothers, father, and uncle. The lone female, Princess, is pregnant and still being studied. After her calf is born and her milk analyzed, she will meet the same fate.

Van Eenennaam worries that this decision, alongside the furor induced by the FDA, will continue to overshadow her positive results. “This should be a good news story,” says Van Eenennaam. Even if she doesn’t agree with the FDA’s approach, the fact that the bacterial DNA was discovered means that the oversight mechanisms for gene editing are generally working, she says. “And the really good news is that this particular edit makes animals not grow horns, it’s passed on faithfully, and the calves are healthy. But that finding hasn’t had the opportunity to be discussed yet.”

Now it will. On Monday, the results of the years’ long study were published with open access in Nature Biotechnology. Whether it will help shape the future of gene-edited animals and a more humane livestock industry, it’s still too soon to say. While most Americans find it unethical to use genetic engineering to make livestock meatier or more nutritious, according to a 2018 Pew poll, public attitudes toward making welfare improvements are less known. On Friday, Van Eenennaam may get a chance to probe those questions, during a Reddit AMA on the science of genome editing. Finally, she’ll be able to discuss it as much as she wants.


More Great WIRED Stories

Read More