An illustration of monkeys on a red and orange background

Cornelia Li

Guoping Feng applied to college the first year that Chinese universities reopened after the Cultural Revolution. It was 1977, and more than a decade's worth of students—5.7 million—sat for the entrance exams. Feng was the only one in his high school to get in. He was assigned—by chance, essentially—to medical school. Like most of his contemporaries with scientific ambitions, he soon set his sights on graduate studies in the United States. "China was really like 30 to 50 years behind," he says. "There was no way to do cutting-edge research." So in 1989, he left for Buffalo, New York, where for the first time he saw snow piled several feet high. He completed his Ph.D. in genetics at the State University of New York at Buffalo.

Feng is short and slim, with a monk-like placidity and a quick smile, and he now holds an endowed chair in neuroscience at MIT, where he focuses on the genetics of brain disorders. His 45-person lab is part of the McGovern Institute for Brain Research, which was established in 2000 with the promise of a $350 million donation, the largest ever received by the university. In short, his lab does not lack for much.

Yet Feng now travels to China several times a year, because there, he can pursue research he has not yet been able to carry out in the United States. In January, I met him in Shenzhen, a city that has gone from fishing village to metropolis during the three decades Feng has lived abroad. He hopped off a red-eye flight from Boston and headed straight to the Shenzhen Institutes of Advanced Technology (SIAT), where he collaborates with several researchers. In front of the institute's headquarters is a large metal sculpture of a motherboard next to a DNA double helix—pairing the technology that defined the 20th century with technology that many think will define the 21st.

Feng had organized a symposium at SIAT, and he was not the only scientist who traveled all the way from the United States to attend: He invited several colleagues as symposium speakers, including a fellow MIT neuroscientist interested in tree shrews, a tiny mammal related to primates and native to southern China, and Chinese-born neuroscientists who study addiction at the University of Pittsburgh and SUNY Upstate Medical University. Like Feng, they had left China in the '80s and '90s, part of a wave of young scientists in search of better opportunities abroad. Also like Feng, they were back in China to pursue a type of cutting-edge research too expensive and too impractical—and maybe too ethically sensitive—in the United States.

At the symposium, the scientists alluded to the potential of using CRISPR, the powerful new gene-editing technique, on the primate brain. The next day, I set out with Yang Zhou, a postdoctoral researcher in Feng's lab, to explore the reality behind that theoretical talk. Our car drove several hours across Guangdong province, passing skyscrapers, then dusty apartment blocks, and then stretches of farmland. During the final half-hour, the GPS instructions petered out. Zhou pulled out his phone to consult photos of road signs he had taken on previous visits.

The car slowed at a leafy turnoff, and Zhou pointed out a sign—brand-new, he remarked—with the breeding facility's name in both English and Chinese. (The company did not want to be named because it feared a backlash from animal-rights activists.) Zhou had been staying there for weeks at a time; the sprawling campus has an on-site cafeteria, and dorms for workers who tend to the facility's thousands of crab-eating macaques. Most of the monkeys are sold to international companies that supply animals to pharmaceutical and research labs.

The breeding facility does not itself genetically engineer monkeys, but Feng realized that its huge number of monkeys made it an ideal proving ground for new genetic-engineering technologies. A Chinese acquaintance was already studying stem cells at the facility, so it was not difficult for Feng and his colleagues to set up shop there, too.

The collaboration between Feng and the facility was spurred by the new gene-editing techniques, especially CRISPR, that have swept like a fever through biology research. CRISPR uses proteins as molecular scissors, allowing scientists to home in on and disable particular genes. Before CRISPR, the genetic engineering of primates was a laborious process capable of a very limited number of edits. Few research groups even attempted it; even fewer succeeded. With CRISPR, monkeys can be genetically engineered almost as easily as mice.

Feng made his career in mice; as an unusually gifted young geneticist, he invented several genetic tricks that advanced the study of rodent brains. When Zhou first joined Feng's MIT lab in 2011, Feng tasked his postdoc with studying autism using mutant mice created in his lab. The mice were "knockouts," in which a particular gene called Shank3 was "knocked out" or disabled. In humans, mutations in Shank3 are found in 1 to 2 percent of cases of autism spectrum disorder, including some of the most severe cases. These patients have the repetitive behaviors and lack of social awareness characteristic of the disorder. They may also be profoundly intellectually disabled and wheelchair-bound.

The knockout mice had characteristics similar to those seen in humans with Shank3 mutations. Certain neurons were underdeveloped, and the mice groomed themselves repetitively—sometimes even ripping open their own skin.

But how applicable are these results to humans? Rodents don't have a full prefrontal cortex—the area of the brain considered the seat of personality, decision-making, and higher cognitive function. And they don't socialize the way humans do. Avoidance of eye contact, for instance, is a classic sign of autism in humans, but eye contact is physically impossible even in healthy mice. "Their eyes are on the side of their head!" says Feng. Autism researchers have become increasingly skeptical of mouse models.

In search of a more humanlike model for his autism research, Feng set finding Chinese collaborators to create Shank3 knockout monkeys. The goal was not to make a monkey with autism, per se, but one with enough symptoms to elucidate the brain structures that cause them and test drugs that could alleviate them. If his Shank3 project worked, Feng wanted to study psychiatric disorders like OCD and schizophrenia in monkeys, too. He told me that a close friend of his had become schizophrenic in college and committed suicide—a tragedy that he kept turning over in his mind. How could things in the brain go so horribly awry? This basic question had animated his research into brain disorders for three decades, and he thought monkeys might finally unlock some of the answers.

China is establishing itself as an international hub of primate research.

At MIT, Feng's lab worked on genetically engineering a monkey species called marmosets, which are very small and genuinely bizarre-looking. They are cheaper to keep due to their size, but they are a relatively new lab animal, and they can be difficult to train on lab tasks. For this reason, Feng also wanted to study Shank3 on macaques in China. Scientists have been cataloging the social behavior of macaques for decades, making it an obvious model for studies of disorders like autism that have a strong social component. Macaques are also more closely related to humans than marmosets, making their brains a better stand-in for those of humans.

The process of genetically engineering a macaque is not trivial, even with the advanced tools of CRISPR. Researchers begin by dosing female monkeys with the same hormones used in human in vitro fertilization. They then collect and fertilize the eggs, and inject the resulting embryos with CRISPR proteins using a long, thin glass needle. Monkey embryos are far more sensitive than mice embryos, and can be affected by small changes in the pH of the injection or the concentration of CRISPR proteins. Only some of the embryos will have the desired mutation, and only some will survive once implanted in surrogate mothers. It takes dozens of eggs to get to just one live monkey, so making even a few knockout monkeys required the support of a large breeding colony.

The first Shank3 macaque was born in 2015. Four more soon followed, bringing the total to five.

To visit his research animals, Feng now has to fly 8,000 miles across 12 time zones. It would be a lot more convenient to carry out his macaque research in the United States, of course, but so far, he has not been able to.

He originally inquired about making Shank3 macaques at the New England Primate Research Center, one of eight national primate research centers then funded by the National Institutes of Health in partnership with a local institution (Harvard Medical School, in this case). The center was conveniently located in Southborough, Massachusetts, just 20 miles west of the MIT campus. But in 2013, Harvard decided to shutter the center.

The decision came as a shock to the research community, and it was widely interpreted as a sign of waning interest in primate research in the United States. While the national primate centers have been important hubs of research on HIV, Zika, Ebola, and other diseases, they have also come under intense public scrutiny. Animal-rights groups like the Humane Society of the United States have sent investigators to work undercover in the labs, and the media has reported on monkey deaths in grisly detail. Harvard officially made its decision to close for "financial" reasons. But the announcement also came after the high-profile deaths of four monkeys from improper handling between 2010 and 2012. The deaths sparked a backlash; demonstrators showed up at the gates. The university gave itself two years to wind down their primate work, officially closing the center in 2015.

"They screwed themselves," Michael Halassa, the MIT neuroscientist who spoke at Feng's symposium, told me in Shenzhen. Wei-Dong Yao, another one of the speakers, chimed in, noting that just two years later CRISPR has created a new wave of interest in primate research. Yao was one of the researchers at Harvard's primate center before it closed; he now runs a lab at SUNY Upstate Medical University that uses genetically engineered mouse and human stem cells, and he had come to Shenzhen to talk about restarting his addiction research on primates.

American scientists worry that the United States is falling behind China on primate research. "I have two big concerns," says Michael Platt, a brain scientist at the University of Pennsylvania who studies primates. "The United States is not investing heavily in these [primate] models. Therefore we won't have the access that scientists have in China." The second, he says, is that "we might lose the talent base and expertise for actually doing primate neuroscience."

China, meanwhile, is establishing itself as an international hub of primate research. While the country does have a burgeoning animal-rights movement, says Peter Li, a China policy specialist with Humane Society International, activists have largely focused on the welfare of pets. Eating dogs has become taboo, and medical experiments on dogs have prompted outrage, but research on monkeys has not faced the same scrutiny.

In the 1990s and 2000s, a number of monkey facilities like the one I visited opened to breed animals for export, mostly to biomedical research projects in the West. This means that China not only has a lot of monkeys but also a lot of experts in monkey reproduction, who can do the delicate experiments required to tinker with monkey genomes.

Cornelia Li

While the U.S. government's biomedical research budget has been largely flat, both national and local governments in China are eager to raise their international scientific profiles, and they are shoveling money into research. A long-rumored, government-sponsored China Brain Project is supposed to give neuroscience research, and primate models in particular, a big funding boost. Chinese scientists may command larger salaries, too: Thanks to funding from the Shenzhen local government, a new principal investigator returning from overseas can get 3 million yuan—almost half a million U.S. dollars—over his or her first five years. China is even finding success in attracting foreign researchers from top U.S. institutions like Yale.

And for American researchers looking to study monkeys in China, every dollar stretches further. A standard monkey in China costs about $1,500, compared to roughly $6,000 in the United States. The daily costs of food and care are an order of magnitude lower as well.

In the past few years, China has seen a miniature explosion of genetic engineering in monkeys. In Kunming, Shanghai, and Guangzhou, scientists have created monkeys engineered to show signs of Parkinson's, Duchenne muscular dystrophy, autism, and more. And Feng's group is not even the only one in China to have created Shank3 monkeys. Another group—a collaboration primarily between researchers at Emory University and scientists in China—has done the same.  

"Under Chinese law, humans are still first."

While in China, I also met Mu-ming Poo, who left UC Berkeley to head up the Institute of Neuroscience in Shanghai. In a few days, scientists in Poo's institute would announce to the world they had successfully cloned monkeys. Such was their national importance that the two cloned monkeys were named Zhongzhong and Huahua after zhonghua, which translates to "Chinese nation" or "Chinese people." Poo was giddy about the breakthrough: With cloning, he said, researchers could more quickly create a colony of identical genetically engineered monkeys instead of engineering one animal at a time. A major challenge with studying monkey models of disease is simply creating enough monkeys to study. Poo imagined a hub in Shanghai that would attract primate researchers from around the world.

Chinese scientists' enthusiasm for CRISPR also extends to studies of humans, which are moving much more quickly, and in some cases under less oversight, than in the West. The first studies to edit human embryos and first clinical trials for cancer therapies using CRISPR have all happened in China.

"China had a reputation of being like the Wild West," says Robert Desimone1, the head of the McGovern Institute for Brain Research at MIT, who first visited the country in 2001. Concerns about fake data, fake peer review, fake research chemicals, and loose ethical standards have all dogged Chinese science. But, Desimone says, "the situation is changing radically." Recent investments in science have drawn Chinese graduate students and postdocs back from the West, and they have brought Western standards with them. Collaborations with American researchers, like the one Feng and Desimone have going at the Shenzhen Institutes of Advanced Technology, have also introduced Western standards to Chinese research institutions.

Feng says his collaborators' research at SIAT was reviewed by a committee similar to the Institutional Animal Care and Use Committees (IACUC), which are required to oversee federally funded animal research in the United States. Though IACUCs are largely made up of scientific experts, they do include local community members, and any committee member can voice ethical objections. "We set up the standards," Feng says—meaning that his group requires its collaborators to meet or exceed U.S. standards.

The breeding facility where the Shank3 monkeys were born is also accredited by the Association for Accreditation of Laboratory Animal Care, an international nonprofit that oversees animal welfare in labs. The standards cover everything from euthanasia practices to the width of hallways in a lab. Desimone and other scientists who had been there gushed to me about the facilities, and by the time our car pulled into the gate I had heard multiple stories about the fruit, toys, and fresh-made steamed buns supplied to the monkeys.

The facility is big enough to require a car to get from one end to the other. (It even has its own sewage treatment system.) So after donning face masks, hairnets, plastic splash shields, gloves, lab coats, and boots, Zhou and I climbed into a white van.

The staff warned us that the monkeys would be excited by new people. Sure enough, as our van pulled up to the breeding houses—several tiled buildings arranged along both sides of the road—dozens of little paws appeared on the bars, and furry heads popped up to get a better look at us. I was wearing a full suit of protective gear, but because I had not been tested for a full battery of diseases, I was told I could only look at the monkeys from afar. A worker wearing a face mask swung open the door to one of the buildings, and curious crab-eating macaques poked their heads out. The species, native to Southeast Asia, has a distinctive head tuft and a long, curling tail. "They're very elegant," said Zhou. One cradled a tiny baby in the crook of her arm.

We moved on to visit the juvenile monkeys, which are housed by the dozen in cages the size of small bedrooms. I saw the much-exalted fresh fruit, and the pastry room where a cook was steaming yams and buns as big as loaves of sourdough.

The tour was a highly managed affair, but facility administrators were still wary of an American journalist. The chairman's assistant, who interrupted the conversation whenever it veered toward what she felt was sensitive territory, emphasized what she saw as the dogmatism of animal-rights activists: "They believe you shouldn't use these animals for experiments, you should protect animals. Human disease, people dying—they won't try to understand these things," she told me. "Under Chinese law, humans are still first."

The Shank3 monkeys born at the breeding facility have since moved to SIAT, where Zhou's collaborators there have been training them in various behavioral tasks. "Monkeys are so smart," Zhou told me, with genuine affection in his voice. When asked the age of his young son, he answered, "Two years old," before quickly adding, as if the two facts are tied in his mind, "The monkeys are three years old now."

SIAT had built a primate lab specifically for the collaboration with the MIT researchers. In February 2014, SIAT promised Desimone that an old student building could be renovated into a primate lab by the end of the summer. "I said, 'That's a joke. That's just a few months away,'" Desimone recalls. He bet a bottle of Maotai, a Chinese liquor,  that they could not make the deadline. The director in Shenzhen bet him two bottles they could. Desimone lost.

"Now it turns out that a lot of things were done very quickly, and they took some shortcuts that in the end, they had to go back and correct," says Desimone. Renovations had to be made for their international accreditation: The floor, for example, could have no seams or lines of grout, which could potentially harbor pathogens. By the time I visited in January, the floor was gray plastic, shiny and smooth. SIAT received their accreditation in early February.

I could not take photos at the monkey lab or breeding facility, and it quickly became obvious why.

On the day I visited, the Shank3 monkeys had been anesthetized for a biopsy, but I went to see the other monkeys housed at SIAT. Crab-eating macaques are given a companion of the same sex. One pair took turns peering me from a landing in their cage. When the lead monkey saw me looking back at them, it would get a little shy and dart behind the other, who assumed the front position. Then it too would become shy, and they would switch, again and again.

The macaque's exquisite sensitivity to faces is one reason Feng wanted to study primates in the first place. When an ordinary monkey sees a picture of an aggressive monkey's face, it will stare right back; it pays less attention to a neutral face, and completely ignores a submissive face. If Shank3 are unable to pick up these social cues—the way some people with autism are unable to identify other humans' facial expressions—it would suggest that primates are indeed a good model for the disorder. To that end, Feng's collaborators in Shenzhen are also studying the monkey's brain with MRIs and EEGs, hoping to identify the ways in which the mutation changes the brain's structure.

I could not take photos at the monkey lab or breeding facility, and it quickly became obvious why. Images are powerful, and I found myself uneasily eyeing the cages and restraint chairs—even as I reminded myself that both are standard lab equipment. The lives of monkeys in captivity suddenly seemed very sad. When I mentioned my reaction to both Feng and Desimone, separately, they gave me the same response: The monkeys in labs are well cared for, and what's more, I shouldn't idealize monkeys in the wild. Wild monkeys get sick; they get eaten; and they fight viciously amongst each other, sometimes to the death. Is that any better, they asked, than living in a lab, supplied with food, shelter, and anesthesia?

A week after my visit to Shenzhen, I met Feng at his office in Cambridge, Massachusetts. Unprompted, he brought up the ethics of genetically engineering captive primates for research. "It's still very early stages," he said during a pause in our conversation. "What should you do? What should you not do? All these things are not finalized yet."

Were there things, I asked, that he thought we should not do? He answered immediately as if the question had been troubling him. Some babies with epilepsy suffer seizures several times a day, he said. "As a parent, you know how painful that is? And you cannot help at all, and you know they're going to die." Feng had almost become a pediatrician after medical school in China. But he could not get used to watching kids die.

Parents with severely epileptic children had asked him if it would be possible to study the condition in a monkey. Feng told them what he thought would be technically possible. "But I also said, 'I'm not sure I want to generate a model like this,'" he recalled. Maybe if there were a drug to control the monkeys' seizures, he said: "I cannot see them seizure all the time."

But is it ethical, he continued, to let these babies die without doing anything? Is it ethical to generate thousands or millions of mutant mice for studies of brain disorders, even when you know they will not elucidate much about human conditions?

Primates should only be used if other models do not work, says Feng, and only if a clear path forward is identified. The first step in his work, he says, is to use the Shank3 monkeys to identify the changes the mutations cause in the brain. Then, researchers might use that information to find targets for drugs, which could be tested in the same monkeys. He's talking with the Oregon National Primate Research Center about carrying out similar work in the United States. "Eventually, we need to do something here," he said, "here" meaning the United States. "We cannot just fall completely behind." Scientists at the California National Primate Research Center have successfully gene-edited primate embryos, but live births have not been reported.

In October, a meeting of experts convened by the National Academy of Medicine in the United States will discuss the implications of editing primate genes. An ethics panel will take up some of the same questions Feng and other researchers are asking themselves: Which diseases are okay to engineer in monkeys? Should monkeys used in research projects be genetically altered to be more humanlike?

One of the speakers on that panel will be Jeffrey Kahn, the director of the Johns Hopkins Berman Institute of Bioethics and chair of the 2011 NAM committee that recommended ending biomedical research using chimpanzees. The National Institutes of Health, in turn, stopped supporting chimp research in 2015. The NIH, which funds the vast majority of biomedical research in the United States, could also effectively limit certain kinds of gene-editing on monkeys by refusing to fund it.

I asked Kahn if strict ethical limits in the United States might allow researchers elsewhere in the world, like China, to outpace researchers here. He noted that it's possible and even completely acceptable that different countries with different cultures might arrive at different limits on this kind of work. "There is a competitive advantage that would be lost if we aren't willing and able to use the technology," he said, "but maybe that's just what we have to accept."

Meanwhile, Poo, a key figure in the China Brain Project, told me, "There's no ethical issues … I don't think there's any hesitation or problem using monkeys as disease models in preclinical trials." As long as the monkeys are well cared for, he said it was no different from the current use of neurotoxins to induce Parkinson's symptoms in monkeys and enable the testing of new treatments.

As primate neuroscience research progresses, scientists will invariably find ways in which monkey models are not perfect—in other words, ways in which monkeys differ from humans. A drug that works in monkeys may fail in human trials. It's when gene-edited monkeys are very good at mimicking human disease—when they are most useful as models—that the ethical questions become most troubling. A scientist in Kunming, China, has proposed using CRISPR to insert human genes related to brain development and language into monkeys. And the same genetic-engineering techniques perfected on monkeys would likely work in humans.

In the scientific literature, monkeys used in research are often described in shorthand as NHP for "nonhuman primate." The distinction is telling in that it needs to be made at all.

  1. On his China visits, Desimone picked up a side gig as a judge on an absurdly popular Chinese game show called The Brain. When I first reached out to him for this story, he said he thought I was contacting him because he's a TV star in China. He gets recognized a lot.