The calf was late. His due date, March 30, had come and gone. At first, Alison Van Eenennaam chalked it up to male calves tending to arrive a day or two on the tardy side. As the week wore on, the animal geneticist reminded herself that gene-edited embryos—like the one that had been growing inside Cow 3113 for the past nine months—can take a little longer to signal to their surrogate mothers that they’re ready to be born. But by the following week, two false alarms at the UC Davis Beef Barn later, with still no signs of impending labor, Van Eenennaam’s fraying nerves had had enough. She called the vet. It was time to induce.
After nearly five years of research, at least half a million dollars, dozens of failed pregnancies, and countless scientific setbacks, Van Eenennaam’s pioneering attempt to create a line of Crispr’d cattle tailored to the needs of the beef industry all came down to this one calf. Who, as luck seemed sure to have it, was about to enter the world in the middle of a global pandemic.
Just weeks prior, California’s governor had ordered the entire state to stay home to avoid spreading a deadly new coronavirus. That was following the discovery of the US’ first case of community spread. The patient was treated at the UC Davis Medical Center, about 20 miles away from the Beef Barn. ICU beds in the Bay Area were filling up. Van Eenennaam was worried about what might happen if the delivery went south and they needed to do a C-section; veterinarians were being asked to save their sedatives to help fill the growing demand for (human) Covid-19 patients on ventilators. And as if that wasn’t ominous enough, the veterinary resident who arrived that day to oversee the birth had spent the morning putting down a number of sheep from the UC Davis herd that had been mangled by coyotes in the night.
“Given how this project has gone, this seriously couldn’t have ended much differently,” Van Eenennaam said, her Australian lilt tinged with uneasy sarcasm. “It’s like the three riders of the apocalypse are probably going to be right on his tail.”
That’s not exactly what happened. The calf arrived that afternoon, 110 pounds and jet black, save for an ankle-grazing splash of white above his rear hooves. Two vets had to extract him from Cow 3113 with chains, but when he was lowered onto the straw-covered barnyard, he was alive and breathing. “Cosmoooooooooo,” Van Eenennaam would shout in triumph. “Welcome to the world, little guy!”
The sky did not darken, and the world did not end. But the black calf, while big and strong and healthy, wasn’t exactly what the scientists had hoped to create. A close look at his DNA would expose just how unpredictable Crispr gene editing can be, and how much more scientists still need to learn before the technology can become routine practice for animal reproduction.
Joey Owen had never really been an animal guy. He’d studied biochemistry and then cancer genetics before bouncing his way into Van Eenennaam’s livestock lab in 2014. It was a hopeful time for scientists like them. Crispr’s genome-engineering potential had been discovered just two years before. It opened up the possibility of creating designer domesticates without the need to port genes from one species into another. Older genetic engineering technologies relied on using viruses and bacteria to shuttle DNA around, triggering an expensive and lengthy approval process from US regulators. As a result, American farmers and ranchers had to this point relied only on the plodding progress of selective breeding to improve the genes of their herds and flocks. Crispr promised to change that.
Van Eenennaam had her sights set on cattle and a gene called SRY—a long stretch on the Y chromosome that instructs mammalian embryos to develop male traits. In nature, there’s an equal chance that cows (and people, for that matter) will give birth to male (XY) or female (XX) offspring. But if she could use Crispr to add a copy of SRY onto the X chromosome of bovine embryos, then she could skew the odds in favor of producing an all-male herd. Any animals with the SRY gene would be physiologically male, even the ones that were genetically XX. That was the hypothesis she wanted to test, anyhow. No one had ever done it before.
In the beef industry, which likes its cattle bigger and meatier, more males means more money means more better. Since male calves gain weight more efficiently than females, farmers could produce the same quantities of meat with fewer animals—potentially reducing the industry’s planet-warming emissions. If Van Eenennaam could prove the concept, that might pave the way for creating other single-sex species—all-female flocks of egg-laying chickens and herds of dairy cows—and ultimately, she felt, a less cruel industrial food production system.
But the experiment she and Owen pitched to the US Department of Agriculture was incredibly ambitious. It would require inserting a very large chunk of DNA into the bovine embryos. When they applied for a biotechnology risk-assessment grant from the USDA, similar edits using Crispr had only been done twice before in cow cells, and never in embryos. Still, the agency awarded them a five-year grant worth $500,000. “When the USDA funded our project we were like, ‘Wait, really?’” recalled Owen. And from the beginning, almost nothing went according to plan.
The first step was trying to get Crispr to make the edits they wanted in bovine embryos. Crispr is essentially a programmable molecular scalpel. It works by targeting a particular location in the genome and slicing through the double helix backbone of the DNA. Then it’s up to the cell’s own repair machinery to stitch the ruptured DNA back together. If you want to add new bits of genetic code between those loose ends, the trick is slipping the cell “template DNA”—in this case a copy of the SRY gene—along with the Crispr components.
This approach works best in cells that are actively dividing. Not so much for single-cell embryos, as Owen soon found out. “We were throwing everything against the wall to see what stuck,” he says. “Nothing did.”
So they started over with an older and less preferred method—editing bovine cells and then cloning their DNA into eggs. (Around the same time, a research group in China was successfully using this technique to Crispr into cattle a gene for tuberculosis resistance.) To make the process easier, Owen added a gene for green fluorescent protein, or GFP, along with the SRY gene. That would allow him to visualize which cells had successfully inserted the genetic recipe for maleness onto the X chromosome. Then he noticed something strange. The cells that did incorporate the new DNA, the ones that glowed, all stopped dividing. The edit had arrested their development.
When they’d begun the experiment, the researchers had targeted a section of the X chromosome that appeared to be within a stretch of junk DNA, far from any life-critical genes. But the only map of the bovine genome that had been available to them at the time was a crude portrait generated more than a decade before. An updated bovine genome released in the spring of 2018 revealed that instead of inserting the SRY gene 10,000 base pairs away from a gene essential for cell growth, as intended, Owen’s team had stuck it right in the middle.
They went back and redesigned their Crispr system, using the updated map to steer their edits clear of any essential genes. Then they tried it with embryos again. This time, it worked. But by now it was the summer of 2018; nearly three years had passed. The project was way behind schedule. They had to ask the USDA for an extension on their grant. The long series of setbacks had Owen feeling pretty despondent, wishing he’d never tried his hand at Crispring cattle. His recent success with editing the embryos in a new X chromosome location had reinvigorated him, but that feeling was short-lived. The first batch of edited embryos they transferred into the uteruses of would-be surrogate heifers didn’t take. Of the next batch, five embryos implanted and made it to the early stages of pregnancy, only to be lost a few weeks later.
Owen and Van Eenennaam consulted with breeders and veterinarians about what they were doing wrong. They suspected that the researchers had damaged the embryos in the lab—perhaps during the biopsy, when they pulled off a tiny bit of the embryo to sequence it and determine if the edit took hold. Doing so takes time, and it requires freezing the embryos until the results are back from the sequencing lab. Each step—the freezing, the biopsy, the editing—decreases the viability of the embryos.
There was a simpler way to do it. They could attach that fluorescence-producing gene again and shine a flash of UV light on the embryos. A green glow would tell them the edit had worked, no biopsy or freezing required. But that would make those animals transgenic; GFP comes from a species of bioluminescent jellyfish that live in the waters off of Washington state. And that would make them genetically modified organisms, or GMOs, subject to the FDA’s arduous approval process. The whole point of the project and using Crispr had been to avoid that.
However, the regulatory landscape had changed while they’d been tinkering. In January, 2017, the FDA decided to classify any edited animal DNA as though it were a new kind of drug. That meant any all-male Crispr herd would be subject to the same regulation as first-generation GMOs. And if, in the eyes of the feds, moving cow DNA around was the same as adding a jellyfish gene, the team figured, why not make their lives a little easier? With little hope that any cattle breeders or commercial entities would be interested enough in their SRY knock-ins to bother tangling with the feds, the researchers might as well go ahead with the glowing gene too.
Van Eenennaam and Owen tried one last time, moving the SRY gene, along with the glowing gene, into about 200 embryos. Since it was their final shot, they decided to make the edit not on the X chromosome, as they had been trying to do, but in a well-established safe harbor site on chromosome 17. Twenty-two embryos survived this process, and of those, nine glimmered under UV light. But only one of them was bright green all over, says Owen. And a month after all the embryos had been transferred into heifers, that bright green one was the only pregnancy that stuck. The research team decided to name the growing calf Cosmo, after a glowing green character in the animated Nickelodeon television series The Fairly OddParents, which aired in the mid-2000s. “I’m obviously too much of a boomer, because I’d never heard of it,” says Van Eenennaam.
Ultrasound suggested that Cosmo was a male. And when he was born on April 7, that was the second thing the vet checked, after making sure the calf was breathing. “Yep, he has testicles—two of ’em!” he told Van Eenennaam and Owen. “Phenotypic male, that’s a good start!”
But to know if he had those male parts because of Crispr would require peering into Cosmo’s DNA. The team drew a few vials of blood from the calf’s neck, and Owen raced it over to the lab and got it into a fridge to start a 16-hour cooldown. He went home, had a few beers to settle his impatient anxiety, and set his alarm for 6:30 the next morning. At 5:00 am, he shot awake and hurried back to the lab while it was still dark. Owen extracted DNA from the calf’s blood and used a technique called PCR gel electrophoresis to look for the presence of the extra SRY and GFP genes. About fours hours later, when the band showed up right where expected, a shock of elation shot through his body. “Holy shit, you really did this!” he thought to himself.
Owen looked around, taking in the moment among the silent instruments. Because of the pandemic, only one person was allowed in the lab at a time; his colleagues were all sheltering at home. So he snapped a pic of the gel and emailed the team.
In her bedroom/home-office, Van Eenennaam clicked on the email, ready for it to be more bad news. Instead, triumph washed over her. “Yessss!” she recalls exclaiming, fists pumping the air.
The gel result wasn’t a total home run. It revealed that Cosmo was XY, meaning he had inherited a copy of SRY from his biological bull-dad, as well as the SRY gene that Owen had Crispr’d onto his 17th chromosome. Even without the editing, he was always going to be a male. But the knock-in had worked, using Crispr in a bovine embryo for the first time ever. “Getting that result was really cool,” says Van Eenennaam. “It was actually one of the best science days ever.”
But there was something else that had shown up in the first, quick scan of Cosmo’s DNA. That was a piece of genetic code that didn’t belong to a cow or a jellyfish, but to a bacteria. To insert such a large gene—SRY is a few thousand DNA letters long—into the single-celled embryo that would become Cosmo, Owen had had to deliver it into the cell the only way scientists knew how: inside a circular piece of bacterial DNA called a plasmid. After Crispr had made its cuts, Cosmo’s repair enzymes had grabbed the plasmid along with the SRY gene and pasted the whole thing into his genome.
This kind of mistake has happened before. A similar plasmid was discovered last year in a pair of genetically dehorned bulls created by Minnesota-based biotech company Recombinetics using the clone-an-edited-cell-into-an-egg approach. Van Eenennaam secured funding from the USDA to study Recombinetics’ bulls and their descendants to see if the genetic alteration was inherited as intended. The plasmid was uncovered by an FDA scientist analyzing DNA sequence data from some of their offspring.
Recombinetics’ scientists had never bothered to look for the presence of plasmid. While such bacterial DNA has not been shown to produce any ill effects on the animals or their meat, in many countries it does redefine them as GMOs, subjecting them to tighter regulatory scrutiny. The revelation scuttled the company’s plan to raise an experimental herd in Brazil from one of the bulls’ semen. It was a huge blow to the small but growing livestock editing industry.
But it succeeded in raising the awareness of such potential problems. So Van Eenennaam’s team had planned ahead of time to do a deep dive into Cosmo’s DNA. That, too, got complicated by the pandemic. UC Davis’s sequencing core was shut down. So the team sent bits of Cosmo’s blood, tissue, and placenta to two different companies to piece together what he was really made of. What they got back was even weirder than they’d expected.
Crispr had made the cuts it was supposed to. But then it made some more. So in the location where Van Eenenaam and Owen had intended to paste a single copy each of SRY and GFP, it got much messier. On one arm of chromosome 17, the new DNA didn’t take at all. The cell randomly grabbed 26 DNA letters to fill the gap. (That’s pretty normal for how cells repair double-stranded DNA breaks.) It was the other arm where the real action happened. In about 90 percent of cells, seven copies of SRY and GFP had been plopped in. Two of them had been inserted backwards. And the bacterial plasmid was in there too. In about 10 percent of cells, there were three (properly oriented) copies of the SRY-GFP construct and one plasmid.
Any way you splice it, that’s a lot of SRY. Plus, there was also the copy on Cosmo’s Y chromosome, the one he’d inherited from his father. “He’s a very manly man,” says Van Eenennaam. These repeats weren't their intention, she says, but so far they don’t seem detrimental to Cosmo. “The fact that he exists tells me that having more copies of SRY than you need doesn’t kill you,” she says.
Studies of mice with extra copies of SRY have not turned up any evidence that the mutations harm the animals, though it can cause sterility in XX individuals. Still, Cosmo’s SRY pile-up is precisely the kind of unforeseen consequence that Crispr critics worry about when it comes to the risks of gene editing. Other experiments abroad, aimed at bringing designer farm animals to market, have turned up strange side effects in recent years, including enlarged tongues in rabbits, pigs with extra vertebrae, and premature deaths of cattle. Lisa Moses, an animal bioethicist at Harvard Medical School, told The Wall Street Journal at the time of those reports, “It’s really hubris of us to assume that we know what we’re doing and that we can predict what kinds of bad things can happen.”
Fyodor Urnov, a gene-editing expert at UC Berkeley’s Innovative Genomics Institute who was not involved in the UC Davis work, says that it’s unrealistic to expect pioneering experiments like Van Eenennaam’s to go flawlessly right off the bat. “They tried to make a bull by targeted integration using gene editing in an embryo, and in fact they succeeded,” he says. Was it a pristine edit? No. Does this mean the field is doomed? Absolutely not, he says. “There are ways to overcome these issues, but you cannot fix them until you know they’re there.”
It’ll take years of close study to see exactly what the effect of all that SRY really is. But the UC Davis team made their initial analyses of Cosmo’s genome public on Thursday morning, at a poster presentation of the American Society of Animal Science. Gaétan Burgio, a geneticist at the Australian National University in Canberra who reviewed the results, says he is not surprised by the less-than-perfect outcome. His group routinely uses Crispr to add new genes to mouse DNA to make new models for studying human diseases. Multiple copies and unwanted plasmid insertions are very typical, he says. “We’ve seen plenty of this in mice,” he says. Sometimes cells will incorporate multiple copies of the bacterial DNA. Burgio has seen up to 70 in one animal. “It’s an absolute nightmare to do knock-ins using Crispr,” he says.
The chaos all comes down to the kinetics of the gene-editing enzymes, he says. Any time you break the DNA cleanly in two, it’s hard to control the outcome. But scientists didn’t really know that in the early years of Crispr gene-editing research, when Van Eenennaam started her all-male cattle project. Reports of erroneous insertions and other undesirable alterations didn’t start showing up in mice until around late 2017, says Burgio.
Some newer Crispr constructs avoid such mistakes by only nicking the DNA, limiting the opportunities for unwanted insertions. But these systems are often fickle in other ways—they might only work in a small fraction of cells or make other kinds of off-script edits. “The bottom line is, there’s no perfect tool,” says Burgio. The design and analysis of the UC Davis team’s cattle experiment look sound to him. Yet it’s further proof that using Crispr to redesign livestock DNA is still a very new science. “Right now we are good at making gene editing in animals efficient, but we are not yet good at making it safe. I think we’ll get there. But we have a fair bit of work still left to do,” he says.
For Van Eenennaam, that work will include raising Cosmo to reproductive maturity, harvesting his semen next spring, and trying to produce a next generation of calves. In theory, he should produce 75 percent male offspring—the normal 50 percent that would inherit his Y chromosome, and an additional 25 percent that would receive the SRY genes Crispr’d into his 17th chromosome. Van Eenennaam hopes to study whether those copies are sufficient to flip on the developmental programming for making chromosomally female animals look and act (and put on weight) like males. She’ll have to apply for a new grant to carry out these next experiments.
It’ll all be just pure research; Cosmo’s jellyfish genes made sure of that, even before the plasmid turned up. Neither he nor his progeny will ever enter the food supply. Their story is likely to end in a UC Davis incinerator.
If the next phase doesn’t work, it may be the end of Van Eenennaam’s pursuit of an all-male beef herd. “Given how hard it is to get a knock-in, I’m not sure I would put my hand up for this project again,” she says.
But that doesn’t mean her efforts won’t help push the field forward. In Australia, researchers are using Crispr to identify the sex of a chicken the day its egg is laid—so that eggs with male embryos can be destroyed before they develop into feeling, cheeping chicks. In Germany, researchers are using Crispr to edit pig sperm so that all their offspring are female, because males grow up to produce an appetite-ruining chemical cocktail called “boar taint” in their meat. Today, farmers often try to avoid that by castrating male piglets without anesthetics, a brutal practice Germany’s government has recently banned. (The law goes into effect next year.) And dozens of other projects aiming to use Crispr to make animals less susceptible to disease and other cruelties of industrial agriculture are in progress in other parts of the world too. Cosmo will have plenty to teach all of them.
Update 7-30-2020 5:30 pm EST: This article was updated to correct that the herd Recombinetics planned to establish in Brazil was for research, not yet for commercial purposes, and to clarify some of the details relating to the FDA’s examination of sequence data from some of the company’s animals.