Just one day after the U.S. surpassed China to become the country with the highest number of Covid-19 cases, the U.S. Defense Intelligence Agency updated its assessment of the origin of the novel coronavirus to reflect that it may have been accidentally released from an infectious diseases lab, Newsweek has learned.
The report, dated March 27 and corroborated by two U.S. officials, reveals that U.S. intelligence revised its January assessment in which it “judged that the outbreak probably occurred naturally” to now include the possibility that the new coronavirus emerged “accidentally” due to “unsafe laboratory practices” in the central Chinese city of Wuhan, where the pathogen was first observed late last year. The classified report, titled “China: Origins of COVID-19 Outbreak Remain Unknown,” ruled out that the disease was genetically engineered or released intentionally as a biological weapon.
“We have no credible evidence to indicate SARS-CoV-2 was released intentionally or was created as a biological weapon,” the report found. “It is very unlikely that researchers or the Chinese government would intentionally release such a dangerous virus, especially within China, without possessing a known and effective vaccine.” Every scientist interviewed by Newsweek for this story also rejected categorically the notion that the virus was intentionally released.
Covid-19 has infected nearly 3 million people across the globe, initially ravaging China before hitting hardest in the West and leaving the United States as the most deeply-afflicted country, with more than 55,000 deaths as of April 27. Its origin remains the subject of not only scientific debate, but a politically charged dispute in the international community.
Citing academic literature, the DIA document states that a “definitive answer may never be known” as to how the disease truly first emerged. A U.S. intelligence spokesperson told Newsweek, “the Intelligence Community has not collectively agreed on any one theory.”
Tracing the origin of a new virus is not easy. It took researchers at the Wuhan Institute more than a decade to trace the 2002-2003 SARS virus to remote bat caves in Yunnan province. It’s not surprising, then, that in early February, China’s Academy for Military Medical Sciences “concluded that it was impossible for them to scientifically determine whether the Covid-19 outbreak was caused naturally or accidentally from a laboratory incident,” according to the DIA document.
Initial assessments conducted by the Chinese government pointed to the city’s Huanan Seafood Market as the likely cause of a natural outbreak of SARS-CoV-2, a new coronavirus that causes Covid-19. In the early days of the outbreak, local officials played down the possibility of human-to-human transmission of the virus and silenced doctors who spoke out about the growing outbreak. It may have undercounted deaths and the number of cases of Covid-19. A spurious theory that the U.S. deliberately planted the virus in Wuhan also started circulating.
China’s foreign ministry told reporters April 23rd that the World Health Organization found “no evidence” the outbreak started at the Wuhan laboratory, and Yuan Zhiming, vice president of the Wuhan Institute of Virology and president of the Chinese Academy of Sciences Wuhan Branch, blasted the inference of intentional misuse or creation as “malicious” and “impossible.”
“The director of the Galveston National Laboratory in the United States made it clear that our laboratory is just as well managed as labs in Europe and the U.S.,” he said. “I think it is understandable for people to make that association. But it is a malicious move to purposefully mislead the people” to think that the virus escaped from [our Wuhan] labs.
“They have no evidence or logic to support their accusations. They are basing it completely on their own speculations.”
The DIA report, however, cites U.S. government and Chinese researchers that found “about 33 percent of the original 41 identified cases did not have direct exposure” to the market. That, along with what’s known of the laboratory’s work in past few years, raised reasonable suspicion that the pandemic may have been caused by a lab error, not the wet market.
Here’s what the scientific and circumstantial evidence shows.
Back in 2002, when SARS emerged in China’s Guandong province, it served as a wake-up call. Over the next few decades, the U.S., China and other nations poured money into efforts to hunt down and catalogue strange new pathogens that live in wild animals and figure out how much of a threat they pose to humans, with the goal of preventing the next devastating pandemic.
In the fall of 2019, the SARS-CoV-2 coronavirus emerged in the middle of the large, cosmopolitan city of Wuhan. Chinese officials at first insisted that the virus, SARS-CoV-2, could be caught only through direct contact with animals. But many of the early patients in Wuhan had no connection to the wild animal markets, which meant that the virus had already been spreading from person to person. When this fact came out, it cast doubt on the veracity of information coming from China, but the virus was well on its way to becoming a deadly pandemic.
In the early days, the prevailing theory of the virus’ origins was that it, like SARS, arose in bats, passed to some other mammal such as a pangolin, and ultimately entered the population through the wild-animal markets.
By March, the wild-virus theory was still the most likely explanation of the origin of SARS-CoV-2–but it was starting to look a little ragged around the edges. For one thing, the Wuhan Institute of Virology, not far from the animal markets in downtown Wuhan, houses the world’s largest collection of coronaviruses from wild bats, including at least one virus that bears a resemblance to SARS-CoV-2. What’s more, Wuhan Institute of Virology scientists have for the past five years been engaged in so-called “gain of function” (GOF) research, which is designed to enhance certain properties of viruses for the purpose of anticipating future pandemics. Gain-of-function techniques have been used to turn viruses into human pathogens capable of causing a global pandemic.
This is no nefarious secret program in an underground military bunker. The Wuhan lab received funding to do this work in part from a ten-year, $200 million international program called PREDICT, funded by the U.S. Agency for International Development and other countries. Similar work, funded in part by the U.S. National Institutes of Health, has been carried out in dozens of labs throughout the world. Some of this research involves taking deadly viruses and enhancing their ability to spread quickly through a population—research that took place over the objections of hundreds of scientists, who have warned for years of the program’s potential to cause a pandemic.
In the years since the SARS outbreak, many instances of mishaps involving the accidental release of pathogens have taken place in labs throughout the world. Hundreds of breaches have occurred in the U.S., including a 2014 release of anthrax from a U.S. government lab that exposed 84 people. The SARS virus escaped from a Beijing lab in 2004, causing four infections and one death. An accidental release is not complicated and doesn’t require malicious intent. All it takes is for a lab worker to get sick, go home for the night, and unwittingly spread the virus to others.
The Wuhan Institute has a record of shoddy practices that could conceivably lead to an accidental release, as officials at the U.S. Embassy in Beijing reportedly warned in a cable on January 19, 2018. “During interactions with scientists at the WIV laboratory, they noted the new lab has a serious shortage of appropriately trained technicians and investigators needed to safely operate this high-containment laboratory,” states the cable, according to the Washington Post.
To be sure, there’s no evidence that SARS-Cov-2 came from the Wuhan lab, nor that the virus is the product of engineering. Most scientists believe, based on the evidence available, that a natural origin is the most likely explanation. But neither have they ruled out these possibilities. “At this stage, it is not possible to determine precisely the source of the virus which caused the COVID-19 pandemic,” says the World Health Organization in a statement to Newsweek. “All available evidence suggests that the virus has a natural animal origin and is not a manipulated or constructed virus.”
The circumstantial evidence is strong enough to warrant putting the lab’s programs and practices at the heart of the investigation. And it’s worth looking anew at whether scientists, in their efforts to protect the public from the threat of natural pathogens, overreached.
Ten years ago, the viral pathogen most in the news was not a coronavirus but influenza—in particular, a strain of flu, designated H5N1, that arose in birds and killed a high proportion of those who were infected. For a while, the virus made headlines. Then it became clear that nearly everyone who caught the bird-flu virus got it directly from handling birds. To cause a plague, it’s not enough that a virus is an efficient killer. It also has to pass easily from one person to the next, a quality called transmissibility.
Around this time, Ron Fouchier, a scientist at Erasmus University in Holland, wondered what it would take for the bird flu virus to mutate into a plague virus. The question was important to the mission of virologists in anticipating human pandemics. If H5N1 were merely one or two steps away from acquiring human transmissibility, the world was in danger: a transmissible form of H5N1 could quickly balloon into a devastating pandemic on the order of the 1918 flu, which killed tens of millions of people.
To answer the question, scientists would have to breed the virus in the lab in cell cultures and see how it mutated. But this kind of work was difficult to carry out and hard to draw conclusions from. How would you know if the end result was transmissible?
The answer that Fouchier came up with was a technique known as “animal passage,” in which he mutated the bird-flu virus by passing it through animals rather than cell cultures. He chose ferrets because they were widely known as a good stand-in for humans—if a virus can jump between ferrets, it is likely also to be able to jump between humans. He would infect one ferret with a bird-flu virus, wait until it got sick, and then remove a sample of the virus that had replicated in the ferret’s body with a swab. As the virus multiplies in the body, it mutates slightly, so the virus that came out of the ferret was slightly different from the one that went into it. Fouchier then proceeded to play a version of telephone: he would take the virus from the first ferret and infect a second, then take the mutated virus from the second ferret and infect a third, and so on.
After passing the virus through 10 ferrets, Fouchier noticed that a ferret in an adjacent cage became ill, even though the two hadn’t come into contact with one another. That showed that the virus was transmissible in ferrets—and, by implication, in humans. Fouchier had succeeded in creating a potential pandemic virus in his lab.
When Fouchier submitted his animal-passage work to the journal Science in 2011, biosecurity officials in the Obama White House, worried that the dangerous pathogen could accidentally leak from Fouchier’s lab, pushed for a moratorium on the research. Fouchier had done his work in BSL-2 labs, which are intended for pathogens such as staph, of moderate severity, rather than BSL-4, which are intended for Ebola and similar viruses. BSL-4 labs have elaborate safeguards—they’re usually separate buildings with their own air circulation systems, airlocks and so forth. In response, the National Institutes of Health issued a moratorium on the research.
What followed was a fierce debate among scientists over the risks versus benefits of the gain-of-function research. Fouchier’s work, wrote Harvard epidemiologist Marc Lipsitch in the journal Nature in 2015, “entails a unique risk that a laboratory accident could spark a pandemic, killing millions.”
Lipsitch and 17 other scientists had formed the Cambridge Working Group in opposition. It issued a statement pointing out that lab accidents involving smallpox, anthrax and bird flu in the U.S. “have been accelerating and have been occurring on average over twice a week.”
“Laboratory creation of highly transmissible, novel strains of dangerous viruses… poses substantially increased risks,” the statement said. “An accidental infection in such a setting could trigger outbreaks that would be difficult or impossible to control. Historically, new strains of influenza, once they establish transmission in the human population, have infected a quarter or more of the world’s population within two years.” More than 200 scientists eventually endorsed the position.
The proponents of gain-of-function research were just as passionate. “We need GOF experiments,” wrote Fouchier in Nature, “to demonstrate causal relationships between genes or mutations and particular biological traits of pathogens. GOF approaches are absolutely essential in infectious disease research.”
The NIH eventually came down on the side of Fouchier and the other proponents. It considered gain-of-function research worth the risk it entailed because it enables scientists to prepare anti-viral medications that could be useful if and when a pandemic occurred.
By the time NIH lifted the moratorium, in 2017, it had granted dozens of exceptions. The PREDICT program, started in 2009, spent $200 million over 10 years, sending virologists all over the world to look for novel viruses and perform gain-of-function research on them. The program’s funding ran out in 2018 and it wasn’t renewed. Early this year, after the Trump administration drew criticism for canceling the program, it granted a six-month extension.
By the time the current pandemic hit, animal-passage experiments had become commonplace. Scientists in many of the more than 30 BSL-4 labs around the world had used them to enhance the transmissibility of respiratory-tract pathogens.
Did the work help during the current pandemic? In a recent article in the Lancet, Colin Carlson, an expert in emerging infectious diseases at Georgetown University, argued that work funded by PREDICT helped virologists rapidly isolate and classify the SARS-CoV-2 virus when it came out. However, the research “could have been better positioned for an overall impact.” Although the program found hundreds of new viruses, it’s nearly impossible for scientists to assess their risk to humans. The only way to tell is to “observe a human infection.”
Richard Ebright, an infectious disease expert at Rutgers, put it more bluntly. “The PREDICT program has produced no results—absolutely no results—that are of use for preventing or combating outbreaks. There’s no information from that project that will contribute in any way, shape or form to addressing the outbreak at hand. The research does not provide information that’s useful for developing antiviral drugs. It does not provide information that’s useful for developing vaccines.”
The Wuhan Institute of Virology is one of many labs to receive PREDICT funding. Shi Zheng-Li, a virologist known as “bat woman” for her group’s work in collecting hundreds of coronaviruses, and her staff at the Institute explored the same bat caves that were thought to have given rise to the original SARS virus in 2002. Her scientists penetrated remote caves, swabbing bats’ anuses and collecting their excretions. When they returned to the lab, they cultured the viruses they found, determined their genomic sequences and tried to determine how they infect cells and animals in the lab.
The Institute began a program of gain-of-function research into bat coronaviruses in 2015. That involved taking selected strains and seeking to increase the ability of those viruses to transmit from one person to another. The gain-of-function research went hand-in-hand with the surveillance project. As scientists identified new classes of bat viruses that have the ability to infect human cells, that raised the question of what changes would have to arise in nature to make that virus transmissible in humans, which would pose a pandemic threat.
In 2015, the Wuhan lab performed a gain of function experiment using cut-and-paste genetic engineering, in which scientists take a natural virus and directly make substitutions in its RNA coding to make it more transmissible. They took a piece of the original SARS virus and inserted a snippet from a SARS-like bat coronavirus, resulting in a virus that is capable of infecting human cells. A natural virus altered with these methods would be easily flagged in a genetic analysis, like a contemporary addition to an old Victorian house.
A virus produced with animal passage methods would be much harder to spot. These viruses are not directly manipulated. When the virus passes from one animal to the next, it undergoes something similar to what would happen in the wild during the course of its evolution. A wild coronavirus passed through 10 ferrets would be difficult to identify as having been engineered or manipulated.
There is no published record of animal-passage work on coronaviruses in the Wuhan Institute. The lab got its first BSL-4 lab in 2018, which is now considered a requirement for this kind of work (though some work proceeds in BSL-3-enhanced labs). It’s possible that researchers started animal passage work in the BSL-4 lab but didn’t finish it in time to publish before the current pandemic, when China tightened up on publications. It’s possible that the work was done in secret. It’s possible that it never happened at all. But some scientists think it’s unlikely that an expensive BSL-4 lab would not be doing animal-passage research, which by 2018 was not unusual.
Tracing the origins
To figure out where SARS-CoV-2 came from, Kristian Andersen of Scripps Research and his colleagues performed a genetic analysis: they published the work, which has been widely cited, on March 17 in Nature Medicine. The researchers focused on certain genetic features of the virus for telltale signs of “manipulation.”
One feature was the spike of protein that the virus uses to attach so effectively to the human body’s ACE2 receptors, a molecular feature of the cells in our lungs and other organs. The spike in SARS-Cov-2, the authors conclude, differs from that of the original SARS virus in ways that suggest it was “most likely the product of natural selection”—in other words, natural, not manipulated in a lab.
However, the paper’s reasoning as to why animal passage, in particular, can be ruled out, is not clear. “In theory, it is possible that SARS-CoV-2 acquired the… mutations during adaptation to passage in cell culture,” the authors write. The theory that the virus mutated in mammalian hosts such as pangolins “provides a much stronger… explanation.” Whether or not that includes animal passage in a lab, they don’t say. Andersen didn’t respond to Newsweek requests for comment.
Rutger’s Ebright, a longtime opponent of gain of function research, says that the Andersen analysis fails to rule out animal-passage as an origin of SARS-CoV-2. “The reasoning is unsound,” he wrote in an email to Newsweek. “They favor the possibility ‘that the virus mutated in an animal host such as a pangolins’ yet, simultaneously, they disfavor the possibility that the virus mutated in ‘animal passage.’ Because the two possibilities are identical, apart from location, one can’t logically favor one and disfavor the other.”
Jonathan Eisen, an evolutionary biologist at UC Davis, says that the preponderance of evidence, while not definitive, suggests that the virus came from nature, not a lab. “There’s no hint there that there’s something unnatural, that is, genetically engineered,” he says. But “there is some wiggle room” in the findings that admits the possibility that the virus was concocted in a lab via animal passage. “Passaging is hard to test for. Escape from a lab is hard to test for,” he says. “If [Wuhan researchers] collected something from the field and they were doing some experiments in the lab with it, and some person got infected and then it spread from there, that would be really hard to distinguish from it having spread in the field directly.”
Wuhan is in possession of a virus, RATG13, that is thought to be the most similar to SARS-CoV-2 of any known virus—the two share 96 percent of their genetic material. That four-percent gap would still be a formidable gap for animal-passage research, says Ralph Baric, a virologist at the University of North Carolina who collaborated with Shi Zheng-Li on the 2015 gain-of-function research. “You keep running into problems that just don’t make it likely,” he says. Wuhan would probably have had to start with a virus closer to SARS-CoV-2 than RATG13, which is within the realm of possibilities.
“The only way to resolve it,” says Baric, “is transparency and open science and have some real investigation into it. I don’t think the Chinese are going to allow that. I don’t know what any country would do in this situation. I would like to think that the U.S. would be transparent.”