A year and a half into the COVID-19 pandemic — and with a death toll approaching 4 million lives — how the coronavirus came to spark a global scourge remains unknown. Was it the result of a spillover from an animal to a human, as has happened repeatedly in the past? Or did the virus accidentally escape from a nearby lab?
The default answer for most scientists has been that the virus, SARS-CoV-2, probably made the jump to humans from bats, if it was a direct spillover — or, more likely, through one or more intermediate mammals. That’s what happened with the coronaviruses responsible for SARS and MERS, and such zoonotic events are standard fare for emerging pathogens.
But without identification of a near-identical virus in a bat or other animal, scientists cannot be completely certain. Stepping into that void is speculation that a naturally occurring or lab-manipulated virus may have inadvertently infected a researcher, who then spread it to others.
In recent months, consideration of the so-called lab leak hypothesis has seemingly gained momentum. In May, a group of 18 scientists wrote a letter in the journal Science criticizing the World Health Organization’s investigation into the virus’ origins, which had ruled lab release “extremely unlikely.” “Theories of accidental release from a lab and zoonotic spillover both remain viable,” the group wrote.
The same month, two former New York Times science journalists penned influential stories backing the lab escape idea, and much of the press has embraced the possibility. President Joe Biden also announced that he was asking U.S. intelligence to issue a new report on the subject by late August.
Despite the increased media attention, little has changed on the ground. There still is no credible evidence that the virus came from a lab in Wuhan, China, where the pandemic began. At the same time, a natural spillover from an animal to a human — the scenario widely viewed as most likely — has not yet been proven.
To some scientists, the lack of evidence about how SARS-CoV-2 emerged has meant little can be concluded either way.
Jesse Bloom, a computational biologist who studies viruses at the Fred Hutchinson Cancer Research Center and the lead author of the letter in Science calling for a more rigorous investigation, told us in an email that he found natural zoonosis and lab accident scenarios involving a researcher being infected with a “natural collected virus” or “experimenting on and possibly growing or modestly modifying a naturally collected virus” all plausible.
“I don’t think there is sufficient evidence to estimate relative probabilities for these scenarios,” he said.
But to many others, the existing data tilts strongly toward a natural spillover.
“[W]hile both lab and natural scenarios are possible, they are not equally likely — precedence, data and other evidence strongly favor natural emergence as a highly likely scientific theory for the emergence of SARS-CoV-2, while the lab leak remains a speculative hypothesis based on conjecture,” Kristian G. Andersen, a professor of immunology and microbiology at Scripps Research, told the New York Times.
“There are still gaps that have to be filled, but I think the evidence we do have right now points to an animal-to-human scenario,” Stephen Goldstein, an evolutionary virologist at the University of Utah who has studied coronaviruses for most of the last decade, told us.
We’ll run through some of the arguments of the lab leak hypothesis and explain why most scientists still suspect a natural origin.
Lab Leak Conjectures
The basic premise of the lab leak hypothesis is that in the course of doing research, a scientist became infected with SARS-CoV-2 and proceeded to spread it to others, kicking off the pandemic.
Theoretically, this could include inadvertent or intentional infection with either a natural virus collected in the field or an engineered or otherwise lab-manipulated virus. (We’ll focus on the accidental scenarios, as these are the ones considered to be more likely.)
As we’ll explain, there is no real evidence for these lab-origin scenarios — and some scientists are adamant the engineered one is a non-starter — but they have not been ruled out.
In support of them, many people question why no intermediate animal has been identified yet and point to the proximity of a top coronavirus lab at the Wuhan Institute of Virology, which is located about a half hour drive from the Huanan Seafood Market that was linked to many of the first COVID-19 cases in December 2019.
WIV is home to a lab headed by Shi Zhengli, a virologist famous for her work tracking down the bat origins of the last coronavirus epidemic. Her lab specializes in collecting coronaviruses in the field and then studying them to understand their potential for creating pandemics.
The lab has made chimeric viruses that mix and match different elements to better understand what’s required to infect human cells — which some people consider to be gain-of-function experiments, although Shi does not. As we’ve written, there is no single definition of gain-of-function, but in this context it typically refers to modifications that aim to make a virus more dangerous or infectious to study potential disease pathways.
Shi told Science that some of her coronavirus research was conducted at biosafety level 2 (BSL-2) — a basic lab safety level that some say is inadequate; this information has also been publicly available in the methods sections of published papers.
Fueling suspicions about the WIV is the institute’s removal of its online database of samples and virus sequences in September 2019 and news of a U.S. intelligence report that three WIV researchers fell ill and sought care in a hospital in November 2019.
Some people also speculate SARS-CoV-2 could have come from an abandoned mine where researchers from Shi’s lab collected bat samples after workers removing bat guano fell ill with an unknown respiratory illness in 2012 and several died. It was a sampling effort there that turned up RaTG13, the bat virus Shi announced in late January 2020 that, at 96.2% similarity, is the closest of all known viruses to SARS-CoV-2 in its overall genome sequence.
A few proponents further argue that the genetic sequence of SARS-CoV-2 contains several unexpected features that are indicative of bioengineering — and that it’s curious that the virus was so well suited to infecting people from the start.
Underlying all of the supposition is China’s lack of transparency and cooperation to find the origin of the virus, which many interpret to be incriminating.
Shi, however, has vigorously denied having the virus or any of its potential precursors and says that no one in the lab has tested positive for the coronavirus, nor do they have antibodies against it. If that’s true, then there’s no way SARS-CoV-2 came from her.
Despite several lab leak narratives that claim RaTG13 could have been modified to create SARS-CoV-2, scientists who study viruses do not believe that’s possible. As others have explained before, RaTG13’s genome differs from SARS-CoV-2 by more than 1,000 nucleotides, making it too different to plausibly have served as a progenitor. “RaTG13 is too divergent to be this ancestral virus,” David Robertson, the head of viral genomics and bioinformatics at the University of Glasgow, told us.
Moreover, Shi says there is only a genome sequence for RaTG13 anyway — live virus was never isolated from the sample — and that she has only ever isolated three SARS-related bat coronaviruses.
The only way SARS-CoV-2 could have come from the lab, manipulated or not, is if the facility was in possession of a virus much more similar to SARS-CoV-2 than RaTG13, multiple experts told us.
“I would estimate at least 99%, that’s the minimum. It probably has to be 99.9% similar to make that kind of switch in the lab at all,” said Robert F. Garry, a virologist at Tulane University School of Medicine. “There’s just no evidence that they had anything close to that.”
In an email interview with the New York Times in June, Shi shot down the notion that she performed risky gain-of-function research, saying her lab had never conducted experiments “that enhance the virulence of viruses.” She also said she had no knowledge of any sick employees at the institute in November 2019, as suggested by a U.S. intelligence report.
In May, the Wall Street Journal reported a few additional details on the timing and number of alleged sick workers from the report, including that they had sought care at a hospital; otherwise, it was the same information in a fact sheet issued by the State Department on Jan. 15, which said the researchers’ symptoms were “consistent with both COVID-19 and common seasonal illness.”
The credibility and significance of the report of illnesses during flu season remains unclear. “No scientist has any way of verifying whether this is true or not,” said Maciej Boni, an associate professor of biology at Pennsylvania State University’s Center for Infectious Disease Dynamics.
In its article, the Wall Street Journal noted that officials have differing views on the strength of the intelligence and that in China, it’s common to visit the hospital for less serious ailments.
If several people did have COVID-19 and were quite ill, Garry said that would mean hundreds of other people would have had COVID-19 at some other level. The workers would also have generated antibodies. “That’s where the seroconversion data comes in,” he said, referring to the antibody tests the WIV says are negative.
Given that Wuhan had a huge surge of flu at the time, he said, it was likely any sick researchers had influenza.
Suspicions have also swirled around the bat viruses Shi’s group collected from a closed copper mine in Mojiang, Yunnan Province, with some opining that the mine workers’ mysterious lethal respiratory disease was COVID-19 or a disease caused by a similar coronavirus and that SARS-CoV-2 could have come from the mine.
A bat sample collected at the mine, which is located in southwestern China about 1,000 miles from Wuhan, in 2013 ended up producing RaTG13, a partial sequence of which made it into a publication in 2016 under a different name. In a November addendum to her 2020 paper describing RaTG13, Shi reported she had collected eight other SARS-related coronaviruses from the mine, which she described in a preprint posted on May 21. All of those viruses are nearly the same and are only a 77.6% match to SARS-CoV-2, falling on a separate branch of the viral family tree than both SARS-CoV-1, the virus that caused the first coronavirus epidemic, and SARS-CoV-2.
Some claim it’s fishy that Shi didn’t disclose the sequences earlier or mention the pneumonia-like illnesses or mine in her publications. Shi explained in her addendum that her lab had tested serum samples from the sick workers for bat SARS-related coronaviruses and they were negative; she retested them more recently for SARS-CoV-2 and they also were negative. She also noted that all of the viruses were only distantly related to SARS-CoV-1 based on an initial check of a single gene sequence; RaTG13 was sequenced nearly in full in 2018 once sequencing technology in her lab had improved and had been renamed from its bat sample ID to “reflect the bat species, the location and the sampling year.” That sequence is what Shi was able to consult following the identification of SARS-CoV-2 in early 2020.
Nothing about the mine story seems abnormal, Garry said, calling it a “distraction.” He added that there would have been no reason for Shi not to report identification of a virus more similar to SARS-CoV-2 than RaTG13 if she had found one.
“We wouldn’t be having this conversation because she would have known which animal or species of bat it came from. And the natural origin would be solved,” he said.
In an interview with Science in July, Shi had already shared the details on the name change and explained that her lab “did not pay special attention” to RaTG13 initially because it was not especially close to SARS-CoV-1.
Edward Holmes, an evolutionary biologist at the University of Sydney, told Science that her explanation made sense. “Of course, they would have been mainly interested in bat viruses closely related to SARS-CoV … not some random bat virus that is more distant,” he said.
As for the missing database, WIV told the WHO team that there had been an online spreadsheet of samples for internal use and there had been plans to make an interactive system, but because of more than 3,000 cyberattacks the data was kept offline.
Lab Escape Possible, But Unlikely
It’s possible, of course, that the Chinese are lying about the database, negative coronavirus test results of WIV staff and that WIV or another facility had SARS-CoV-2 or its precursor in one of its labs.
Lab escapes of dangerous pathogens have happened in the past, including multiple instances with SARS-CoV-1 in China. It’s nevertheless the case that there has never been a lab accident that sparked a pandemic or led to an outbreak of a novel pathogen, nor has there ever been a known breach at the WIV. (One incident frequently cited as an example of a lab accident, the 1977 influenza epidemic in Russia, is likely to have been a vaccine trial gone awry, not a lab release.)
The WHO team was not able to independently verify the lab’s virus collection or safety records, although that was never the mandate for the organization’s origins studies.
Some scientists would nevertheless like a more comprehensive investigation.
Ralph Baric, an epidemiologist and coronavirus researcher at the University of North Carolina at Chapel Hill who previously collaborated with Shi and was a signatory on the Science letter criticizing the WHO, told us in a statement that the genetic structure of SARS-CoV-2 “points to the virus originating in natural wildlife populations, most likely bats, that passed from animals to humans” but that “more investigation and transparency are necessary to define the origin of the pandemic.”
“For example, a rigorous investigation would have reviewed the biosafety level under which bat coronavirus research was conducted at WIV,” he added. “It would have included detailed information on the training procedures with records, the safety procedures with records and strategies that were in place to prevent inadvertent or accidental escape.”
Dr. W. Ian Lipkin, an epidemiologist at Columbia University and a co-author on an influential Nature Medicine paper from March 2020 that found it “improbable” that SARS-CoV-2 was manipulated in the lab, has also raised concerns about safety. He told former New York Times science reporter Donald McNeil that he was troubled by the fact that some of Shi’s coronavirus work had been done in BSL-2 labs.
Despite the story’s suggestion that Lipkin might no longer agree with the Nature Medicine paper’s conclusion that the virus was not bioengineered, he confirmed to FactCheck.org that he did.
“I said only that novel bat viruses should not propagated at BSL-2 and that this raised concerns about biosecurity at the WIV,” he said in an email. “I don’t disavow the paper.”
He has since told the Washington Post that it’s possible that WIV researchers might have unwittingly become infected with a coronavirus they hadn’t yet characterized.
Possible, however, does not mean equally likely, as Lipkin readily acknowledged. And indeed, even some who signed the Science letter, such as Baric, think the most probable scenario is a natural one.
Seeing how the letter has been interpreted to support the lab leak hypothesis, one signatory, Caltech professor of biology and bioengineering Pamela J. Bjorkman, has backed away from it.
“I thought the letter would have the effect of prompting more funding for searching for natural viruses in animal reservoirs, which I personally have always assumed represent the origin of SARS-CoV-2 infections in humans,” she wrote in a letter to the podcast This Week in Virology. “Perhaps naively, I did not anticipate that the letter would be used to promote the lab origin hypothesis.”
Marc Lipsitch, a professor of epidemiology at the Harvard T.H. Chan School of Public Health and another co-author of the Science letter, also told Vice that the letter has been misinterpreted as backing the lab leak hypothesis instead of being a call for more research. He has emphasized the lack of evidence on the subject and told CNN that the lab leak hypothesis “is not a fringe theory” and should be investigated.
On the flip side, there are a few scientists who have reversed their opinions and now think a lab origin is more likely than a natural one.
But many scientists — especially the ones with the most expertise in coronaviruses — find a lab origin unlikely, even if they cannot exclude it.
“The finding of SARS-CoV-2-like viruses circulating in horseshoe bats in both China and Southeast Asia, coupled with the strong links of the first cases to animal markets in Wuhan are very compelling evidence SARS-CoV-2 is the result of an animal associated spillover much like SARS,” said Robertson, the University of Glasgow virus bioinformatician, who has studied how SARS-CoV-2 might have evolved. “On lab-leak, there’s no evidence that SARS-CoV-2 escaped from a lab other than the coincidence of the Wuhan Institute of Virology being there.”
Goldstein, the University of Utah coronavirus virologist, agreed.
“We know that a majority of the first cases that were picked up were directly linked to animal markets in Wuhan. We know that these SARS-related coronaviruses circulate in animals; we know that people are infected with these viruses,” he told us, citing a study Shi’s group did that found 2.7% of people in a rural village had antibodies against bat SARS-like viruses, indicating past infection.
Additionally, he said, a study published in Scientific Reports in June showed the presence of multiple mammal species illegally being sold in wet markets across Wuhan between May 2017 and November 2019, adding to the plausibility of how viral transfer might have occurred.
“And so, kind of all the ingredients are there, the epidemiological links are there; the scientific evidence for the virus being engineered doesn’t hold up at all,” he said.
Garry, the Tulane virologist, also noted the past examples of natural spillovers, including SARS; the fact that the first four known COVID-19 cases in Wuhan had links to different wet markets, as shown in the WHO report — and no sign that Shi’s lab had any virus close to SARS-CoV-2.
“There is no evidence at all for a lab leak. Nothing scientific, it’s just an accusation,” Garry said. “You have to think one of the leading virologists on the planet is part of a major conspiracy that is involving hundreds of people.”
Viral Genome Almost Certainly Not Engineered
Early in the pandemic we repeatedly debunked baseless conspiracy theories circulating on social media about SARS-CoV-2 being bioengineered.
For example, there were bogus claims that the virus contains HIV “insertions” and false claims that the virus was created by a prominent Harvard chemist who was charged by the Department of Justice on Jan. 28, 2020, for making false statements about his ties to China.
Many scientists remain open to a lab escape of a natural virus, but fewer entertain the notion that SARS-CoV-2 was engineered. While this cannot be ruled out entirely, multiple coronavirus experts view this as implausible.
“I am completely confident that the virus was not engineered,” University of Pennsylvania coronavirus researcher Susan Weiss told us in an email.
University of Utah’s Goldstein said it was “virtually impossible,” while Dr. Stanley Perlman, a coronavirus researcher at the University of Iowa, went with “impossible.”
In March, a group of scientists, including Kristian Andersen of Scripps Research and Garry of Tulane, published a paper in Nature Medicine that combed through the genome sequence for any signs of lab tinkering and concluded there were none.
Initially, the researchers had been suspicious that there were elements that were engineered. But upon closer examination, the group rejected that hypothesis, as we have written. (Contrary to some suggestions, the Andersen paper was not just an opinion piece that had no vetting by other scientists. A spokesperson for Nature Medicine told FactCheck.org by email that the paper was peer-reviewed.)
Even if scientists used methods that would not leave a trace of manipulation, as some lab leak proponents have suggested, that would still leave the arguably insurmountable problem of not knowing enough to create the virus.
“No one would know how to do it,” Perlman said. “If one doesn’t have the virus in hand, how do you decide to make this?”
Recently, there has been additional speculation about SARS-CoV-2’s furin cleavage site, which is a spot on the virus’s spike protein that’s cut by the enzyme furin to activate the spike and prepare the virus for entering cells. Experiments have shown the site is required for the virus to infect human lung cells and for viral transmission in ferrets. At first glance, the site is potentially curious, as it’s absent in coronaviruses that are closely related to SARS-CoV-2.
Furin cleavage sites, however, exist in many other coronaviruses, such as feline coronaviruses and the virus that causes MERS. Because similar sequences for the cleavage site are found in other coronaviruses, “its presence is not at all suspicious or indicative of lab manipulation,” Robertson said.
“The lineage SARS-CoV-2 emerged from is under-sampled so it’s not surprising there’s some unique properties in its genome,” he added.
Thomas Gallagher, a professor of microbiology and immunology at Loyola University Chicago who studies coronaviruses, also said he did not think the furin cleavage site was a sign of engineering.
“Some coronaviruses naturally have furin cleavage sites, others do not,” he told us in an email. “These cleavage sites evolve naturally under various natural selective pressures. The selective pressures are often powerful, so the furin cleavage site is a hotspot for coronavirus variation.”
In a self-published story on Medium, later posted on the Bulletin of the Atomic Scientists’ website, former New York Times journalist Nicholas Wade zeroed in on an ostensibly suspicious element of the furin cleavage site. Namely, that the underlying genetic sequence of the virus’s cleavage site looked manipulated because of two CGG stretches that code for the amino acid arginine. Because CGG is not often found in coronaviruses, he argued, instead of evolving naturally, it was more likely that a scientist had gone in and inserted the site into the genome while doing gain-of-function research.
In support of his theory, Wade quoted David Baltimore, Nobel laureate and president emeritus of CalTech, as saying the furin cleavage site with its arginine codons was the “smoking gun for the origin of the virus” and that it made “a powerful challenge to the idea of a natural origin for SARS2.”
But on Twitter, Andersen pushed back, noting that while rare, CGG triplets are not unheard of in SARS-CoV-2’s genetic sequence and are used to code for arginine 3% of the time. Indeed, some feline coronaviruses differ there from SARS-CoV-2 by just one nucleotide. And now that the world is awash in SARS-CoV-2 sequences, there’s no sign of the virus mutating away from using those triplets at the cleavage site, which might be expected if the sequence was unnatural.
Informed of Andersen’s points, Baltimore told a journalist with Nature that he agreed that the site could have evolved naturally. FactCheck.org contacted Baltimore as well and in an email he acknowledged he “shouldn’t have used the phrase ‘smoking gun’ because it sounds so definitive,” although he added that he didn’t think Andersen “is giving enough credence to the possibility that the furin cleavage site had a non-natural origin.”
Virologists, however, say there are plenty of other reasons why it’s incredibly unlikely that the furin cleavage site was engineered, starting with the fact that the site is not a very good cleavage site.
“This is a pretty bad one; it’s not cleaved very efficiently by furin,” Goldstein said.
In fact, he said that based on other coronaviruses with similar cleavage sites, it’s known that mutations that make the protein sequence closer to the SARS-CoV-2 sequence end up losing the ability to be cut.
“If you’re trying to insert a furin cleavage site,” Goldstein said, “why would you pick a furin cleavage site that is not actually a functional furin cleavage site in other viruses?”
Additionally, the cleavage site exists as an insertion in the genome that strangely breaks up the triplets in what is called an “out-of-frame” insertion. Any scientist wanting to add a furin cleavage site “would just plop it in nice and clean,” Goldstein said. “I don’t how to explain from a scientific standpoint how ridiculous this is, the idea that you would do an out-of-frame insertion. It just makes no sense.”
Garry, the Tulane virologist, was also baffled by the suggestion that the cleavage site sequence showed the virus had been engineered. “Which graduate student or post doc would think to put it in out-of-frame? That part I just don’t get,” he said. “This, for all the world, looks like a natural virus.”
Another line of speculation is that instead of a scientist deliberately choosing what to modify, the virus was serially passaged through human cells or an animal. That, in theory, would eliminate the requirement for a scientist to know what to insert or change. Lab leak proponents often cite experiments with human cells or humanized mice as a potential way this could happen.
But Perlman, who has done experiments passaging coronaviruses in mice, said that would not work. “Most of the time when you take viruses and pass them in tissue culture cells, you get cells that grow very well in tissue culture cells and nowhere else,” he said. And humanized mice are still mostly mice, he said, so the virus would adapt to growing better in mice, not humans.
“It’d have to be something nearer to a palm civet cat, which is a weird animal to be passaging it [the virus],” Perlman explained.
You’d also need a starting virus that is much closer to SARS-CoV-2 than any known virus, he said, and even then, the virus you’d end up with would almost certainly not be SARS-CoV-2.
As a result, Perlman said, such a scenario could be technically possible but is extraordinarily improbable. In his mind, the engineering scenario can be ruled out, although he still considered accidental release of a natural virus as an unlikely, but possible, pathway.
Further complicating the lab leak scenarios is that when SARS-CoV-2 is grown in the standard cells used to isolate and propagate viruses in the lab, the furin cleavage site is frequently lost, as is documented in multiple reports. The Shi lab, notably, used those cells with each of the three SARS-related bat coronaviruses it successfully isolated in the past.
Some have also argued that SARS-CoV-2 was too well adapted to infecting humans at the start of the pandemic — and that this could indicate human design.
But Penn State’s Boni said that’s a faulty line of thinking.
“There’s no guarantee that something that crosses over has to be perfectly adapted or half adapted or a third adapted. Whatever happens, happens,” he said.
The H1N1 swine flu pandemic in 2009, for example, he said, was very well adapted to humans and took off very easily and very quickly. “It is not a sign that they were bioengineered,” Boni said.
A paper he co-authored with Robertson in PLOS Biology pieced together SARS-CoV-2’s evolutionary history and suggests that the virus’s ability to infect a broad range of mammals evolved hundreds of years ago.
“This would indicate that the SARS-CoV-2 progenitor did not have to adapt to humans much, if at all,” Robertson said, because it had already become a “generalist virus” long ago, although he said an intermediate animal could very well still be involved in the transfer to a human.
For some, it’s more than a coincidence that a novel coronavirus outbreak began in the same place as China’s premier coronavirus research lab.
But Wuhan is also a city of 11 million people and a hub for commerce, including the wildlife trade.
“There’s thousands and thousands of large and small markets in a city like Wuhan where there’s human-animal contacts every day,” said Boni, who spent eight years doing field epidemiology in Vietnam. “These human-animal contacts aren’t rare. People don’t do their shopping at Whole Foods, people do their shopping at these markets.”
Thus, while lab accidents do occur — and Boni said the possibility should be investigated — it “doesn’t really compare to the scale of human-animal contact that you have in a province like Hubei where there’s 60 million people and on an average day 5 million of those people could have been in contact with an animal at a market.”
Lacking more information, Boni said he thought a natural spillover for SARS-CoV-2 was “a thousand times, a million times more likely” than a lab leak.
Lack of a Clear Animal Connection
Still, it’s true that there is no proof of a natural spillover, and some 18 months out from the first identified COVID-19 cases, the lack of an animal that transmitted the virus to humans has led some people to wonder whether there was one.
After all, with the first SARS epidemic in 2003, cat-like mammals known as palm civets were identified as possible intermediate hosts within several months and fingered more definitively within a year. And with MERS in 2012, it took about a year to find out that people had likely picked up the virus from camels.
But experts told us the delay is not unexpected.
“It’s not really surprising,” said Goldstein. For one, unlike with the first SARS, the market linked to many of the early COVID-19 cases was quickly shut down, making it significantly more difficult to find any potential intermediate animals there.
“You need to get lucky,” he said. “You have to go at the right time. If you go later, it’s going to be hard.”
And, as Perlman, pointed out, “If I were illegally trading in exotic animals and heard [a] SARS-CoV-2 pandemic was about to begin, the first thing I’d do is take my exotic animals and high-tail it.”
“In China, it is not surprising that scientists did not find SARS-CoV-2 in potential animal sources immediately after the human outbreak in Wuhan. Nor does that result indicate there is a problem with the wildlife spillover theory,” wrote Christine K. Johnson, director of the EpiCenter for Disease Dynamics at the One Health Institute at the University of California, Davis’ School of Veterinary Medicine, in an editorial in Scientific American. “This is a difficult search that takes time.”
There have been some efforts to look for an intermediate. The WHO reported that 80,000 wildlife and farm animal samples from China had been tested, all of which turned up negative for SARS-CoV-2. But Garry said that the figure is not as impressive as it might seem.
“80,000 animals sounds like a lot, but a lot of those were domestic cattle and chickens and birds and things like that that wouldn’t be expected to have SARS-CoV-2,” he said. “When you actually get down to the species that might have it, it’s maybe a few hundred samples at most.”
And there may not be an intermediate. Either way, identifying a closer virus in bats, which are thought to be the original source of the virus, could also take time.
With SARS, it took until 2017 to find a population of horseshoe bats harboring viruses with all the specific features of SARS-CoV-1 in a cave in Yunnan, demonstrating that the virus likely originated in those animals and then likely passed through civets to infect humans. That detective work, of course, was done by Shi.
“It’s taken a long time to figure out a lot of these zoonotic infections and I suspect it will take a few years more,” Perlman said of SARS-CoV-2.
Already, scientists have identified multiple bat viruses that are closely related to SARS-CoV-2, including RaTG13, although no precursor virus has been found.
Notably, groups other than Shi’s have also found three other bat viruses, all from Yunnan, that are as much as 94.5% identical to SARS-CoV-2. When factoring in viral recombination, or the viral habit of exchanging chunks of genetic material, these are even more closely related to SARS-CoV-2 than RaTG13.
The existence of these sequences in nature, many scientists say, bolsters the case for a natural origin, even if it doesn’t prove it.
“A reasonable speculation is that additional animal sampling could identify sarbecoviruses [SARS-related coronaviruses] that are even closer to SARS-CoV-2, even some that have the furin cleavage site,” said Gallagher, adding that these discoveries thus far offer “a data-based reason why the natural spillover scenario is favored.”
It’s also possible a conclusive animal connection will prove elusive.
“We may not get a definitive answer,” Goldstein said. “Which unfortunately is not unusual in science. I mean, we still don’t definitely know the host of Ebola virus, which we’ve been looking for since 1976.”
Getting an Answer
Indeed, while several experts are optimistic that more sampling will turn up more evidence, the full pathway for how SARS-CoV-2 made it into people might still have some gaps.
“As more sampling is [done] on horseshoe bats, there’s a very good chance we’ll find closer animal viruses to SARS-CoV-2,” Robertson said. “While we might never know the exact route of transmission we should be able to understand the process that occurred.”
If an animal reservoir is found, Boni said, it’ll be pretty strong evidence that SARS-CoV-2 did come from that animal reservoir. “But are we ever going to have definitive, categorical proof that it wasn’t sampled by a researcher and then mishandled and accidentally leaked? We may never have that direct, 100% proof,” he said.
In the clearest case, researchers would be able to identify intermediate animals in proximity to a bat population carrying viruses similar to SARS-CoV-2 and a minority of people living in the area would also have antibodies to the virus, Boni said.
To some, that still might not be definitive proof that the virus didn’t leak from a lab, but it’d be very hard to argue with.
It’s also possible evidence will surface implicating a lab. In his interview with the New York Times, Scripps’ Kristian Andersen said new evidence could emerge that could shift his opinion to favor the lab leak hypothesis. “For example, any credible evidence of SARS-CoV-2 having been at the Wuhan Institute of Virology prior to the pandemic — whether in a freezer, in tissue culture or in animals, or epidemiological evidence of very early confirmed Covid-19 cases associated with the institute,” he said.
Gallagher also said that while release of a natural virus “seems very unlikely,” he wasn’t familiar enough with the operations of the Shi lab to comment further.
“I remain open to new findings as they arise,” he said. “There are still many unknowns and quite obviously, new discoveries can influence my positions.”
As it stands, though, some scientists, while still supportive of further investigation, say there is little to no reason to suspect a lab is the source of SARS-CoV-2 — and focusing too much on the possibility is diminishing the chances of finding out what happened.
“If we’re going to get the answer, we have to do it with some degree of diplomacy,” Dr. Anthony Fauci, the director of the National Institute of Allergy and Infectious Diseases, said in a podcast with the New York Times. “Because if we want to be part of the team that goes out there and finds out is there a connection with an animal that might have been brought in for many, many, many miles away into the Wuhan markets, we’re going to have to do that in collaboration with the Chinese.”
Editor’s note: SciCheck’s COVID-19/Vaccination Project is made possible by a grant from the Robert Wood Johnson Foundation. The foundation has no control over our editorial decisions, and the views expressed in our articles do not necessarily reflect the views of the foundation. The goal of the project is to increase exposure to accurate information about COVID-19 and vaccines, while decreasing the impact of misinformation.