Of the many mysteries kindled by the coronavirus pandemic, the question of where SARS-CoV-2 originated has been the most difficult to answer.
No matter how history ultimately writes the pandemic's origin story, it was, almost certainly, an unfortunate accident. But what kind of accident? That urgent question is key to preventing the emergence of a SARS-CoV-3 or a COVID-29, but an uneasy tension has been building around the answer. Two conflicting narratives have materialized since the first cases were detected in the central Chinese city of Wuhan a year ago, exposing a deep chasm between researchers exploring COVID-19's beginnings.
Was it a natural accident? A bat coronavirus found its way into human lungs sometime in recent history. It evaded our defenses until chance mutations made it more virulent and deadly than any coronavirus before it. This accounting is backed by scientific data and a detailed history of viruses jumping the species barrier.
Or was it a laboratory accident? A bat coronavirus found its way out of the Wuhan Institute of Virology, a high-security facility in the heart of the city, and snuck into the population. This theory is the most uncomfortable and most controversial. If true, it would have severe and lasting ramifications on research, geopolitics and trust in scientific institutions. "This keeps us up at night," says Stuart Turville, an immunovirologist at the Kirby Institute in Australia. "This is the nightmare within nightmares."
Many scientists still consider a natural origin to be the most likely starting point. Yet, just as the World Health Organization's task force descends on China to investigate COVID-19's emergence and mainstream publications like New York Magazine propagate fanciful versions of what could have happened in Wuhan, the "lab leak theory" has wended its way back into the spotlight.
Over the last year, that theory has become increasingly difficult to ignore. Coincidences and circumstantial evidence continue to build, pointing to the Wuhan institute as a potential starting point. But the theory, and a dearth of information, has also helped spawn baseless conspiracies, like the notions that COVID-19 is a bioweapon or that it was used as a cover to install 5G across the world.
This tangled web of conspiracy and politicking has often seen those who support investigating the lab leak treated with contempt, their theories dismissed offhand. Heated exchanges and toxic feuds have flared between scientists online and in the press. Researchers have been harassed, abused and threatened. Many have been reluctant to speak out at all.
Investigating a possible accidental lab leak is a valid line of inquiry. Proponents of the lab leak theory argue that without a full accounting of the work performed at the WIV we may never truly know where the coronavirus came from -- and that could stifle our efforts to prevent the emergence of the next pandemic. To understand why a lab leak is so plausible to so many, we must stitch together clues, stretching a decade into the past.
The first lies in a damp cave, hidden in the southern corner of China.
Pandemics begin in medias res. Scientists and epidemiologists are thrust into action and have to work backward to determine where and how a new pathogen first jumped to humans. It's a complicated process involving ecologists, epidemiologists, geneticists, virologists and a legion of expert investigators. "It typically takes years to find reservoir hosts -- if we even find one at all," says Kristian Andersen, a virologist at Scripps Research Institute in San Diego.
Scientists agree on one immutable fact: The closest known relative of SARS-CoV-2, the coronavirus that causes COVID-19, was discovered in 2013 in bat fecal samples obtained from an abandoned mineshaft in Yunnan province, China, about 1,000 miles southwest of Wuhan. A year earlier, the shaft was ground zero for a perplexing spate of mystery illnesses.
On April 2, 2012, a 42-year-old resident of Yunnan province known only as "Lu" descended into the derelict mineshaft cut into the side of a hill near the remote village of Tongguan. For two weeks, he scrubbed walls inside the mine, battling against the rancid stench of bat feces. By the time he'd completed his shift, he was running a high fever and, occasionally, coughing up rust-colored mucus and clots of blood. On April 25, he was admitted to the First Affiliated Hospital of Kunming Medical University.
Within 10 days, five colleagues who'd cleaned the mine with Lu were also admitted, showing symptoms eerily familiar to Chinese doctors. Chest X-rays showed their lungs filled with fluid. Analysis of their blood found immune cells depleted.
The creeping shadow of SARS stretched all across the miners' cases. A decade earlier, China had lived through an epidemic caused by the SARS coronavirus that killed 774 people between 2002 and 2003. Doctors at the Kunming hospital consulted with experts from across China, including Zhong Nanshan, a respiratory specialist instrumental in developing treatments for SARS patients. Nanshan suspected the patients may have picked up a SARS-like virus and recommended they be swabbed and tested for antibodies.
Three of the six patients, including Lu, died within weeks. Weighing the evidence from a battery of tests, doctors concluded the mine cleaners were likely infected with an unknown virus while working in the shaft, possibly due to their close proximity to bats and guano. The mine was closed, and its residents -- including rats, shrews and Chinese horseshoe bats -- became major suspects in the miners' deaths. An investigation began.
Shi Zhengli, a virologist at the WIV, and her team were called in to search for viruses around the mine. Shi had, in 2005, helped identify bats as reservoirs of potentially deadly coronaviruses. She became known as China's "bat woman." Between 2012 and 2015, her team laid huge plastic sheets underneath bat roosts, collecting the feces that accumulated there and shipping the samples around the world. Some were delivered to the WIV, others went to the Australian Animal Health Laboratory in Geelong, a city south of Melbourne.
At both locations, the samples were screened for pieces of genetic material from a range of infectious agents, including coronaviruses. One sample, dubbed "BtCoV4991," looked similar to the virus that caused SARS in 2002. It was entered into an online database known as GenBank in 2016 and, for the most part, it was forgotten about.
But when SARS-CoV-2 emerged in Wuhan in December 2019, the miners and BtCoV4991 were thrust back into the spotlight.
In the first paper detailing the novel coronavirus in February 2020, Shi's team compared it with another coronavirus: RaTG13. This virus has become one of the most critical elements of the origin story, and both sides of the debate latched onto it.
The genetic sequence of RaTG13 shares 96.2% of its genome with SARS-CoV-2, making it a distant relative of the virus, one that can't infect humans. Further examination of this viral genome and BtCoV4991 showed they were actually the same virus. That put RaTG13 at the scene of the Mojiang cave, lurking in the abandoned mineshaft within the bodies of horseshoe bats.
RaTG13 left a loose thread. The illness that killed the three miners was similar to COVID-19, according to a 2013 master's thesis from a student at China's Kunming Hospital. Could the miners have been infected with the novel coronavirus that has now spread across the globe? How about one of its relatives? A reanalysis of stored blood samples from the miners by Shi's group did not show evidence they were infected with either a SARS-like virus or SARS-CoV-2.
For now, we can't say what killed them -- and that absence of a clear answer is where some of the more imaginative conspiracies have flourished.
Proponents of the lab leak theory suggest the Mojiang mine incident provides a reasonable path for a virus similar to SARS-CoV-2 to be brought to Wuhan. If RaTG13 was found in the caves and brought back to be used in experiments at the WIV over the last seven years, some say, perhaps it was able to pick up mutations in "gain of function" experiments, in which researchers tinker with a virus' genes to study how they interact with human cells. While these types of experiments did take place at the WIV, there's no evidence RaTG13 itself has been manipulated in the lab. It's possible, but requires further investigation of the institute's records.
However, RaTG13 is unlikely to have been used, Roger Frutos says, because while it is real, viral particles were not isolated from animals. All researchers have is chunks of the virus' genetic code. "This virus exists only as a virtual sequence in the computer," says Frutos, a molecular microbiologist at the French Agricultural Research Centre for International Development, or CIRAD. "It's an avatar," he notes.
The Mojiang mine remains an interesting thread. Journalists from the Associated Press and the BBC have since attempted to visit them and, in both instances, were tailed by Chinese authorities and denied access to the area. Authorities have also confiscated samples from researchers who were permitted into the caves, according to AP. The tightly controlled reporting from the country and lack of transparency from Chinese scientists has raised skeptics' eyebrows even further.
Some have even proposed that RaTG13 data, generated by researchers at the WIV, is "fake" and that its records were produced as a coverup. There is no evidence for this.
But the fact that relatives of the SARS-CoV-2 virus were discovered so far from Wuhan and then researched at the WIV has placed constant scrutiny on Shi Zhengli's work and the safety protocols in Chinese laboratories.
In laboratories around the world, viruses are contained within specialized, high-security facilities operating under strict safety standards, like supermax prisons. When a virus breaches containment, it's "leaked." But it's more like an escape. Viruses will take any opportunity to make a jailbreak.
On a few occasions, they've been able to slip their shackles and sneak out.
In 2004, two lab workers at the National Institute of Virology in Beijing became ill with pneumonia. They had inadvertently been infected with the SARS coronavirus after "two separate breaches of bio-safety," according to the WHO. The accident resulted in 11 cases and one death, only a year after the SARS outbreak had been contained.
"The second, third, fourth and fifth entries of the original SARS coronavirus into human populations occurred as a laboratory accident," says Richard Ebright, a chemical biologist at Rutgers University who has long had concerns about the safe use of high-level biosafety laboratories.
For those who believe the novel coronavirus leaked from the WIV, incidents like this are seen as damning evidence that leaks are not only possible, but likely. A researcher, working on a virus related to SARS-CoV-2, is accidentally infected, leaves the lab and begins silently spreading the virus through the community.
On Jan. 15, outgoing US Secretary of State Mike Pompeo released a statement suggesting the US had "reason to believe" researchers at the WIV became sick in autumn 2019 with "symptoms consistent with both COVID-19 and common seasonal illnesses." (The State Department did not respond to a request for comment.) Could a leak have happened again?
Shi considered this possibility when she first heard about a new coronavirus spreading in Wuhan, according to an interview given to Scientific American on March 11. Other researchers, too, have contemplated such a scenario.
On March 17, 2020, Scripps' Andersen and four other esteemed virologists co-authored a letter to the editor of the journal Nature examining features of the SARS-CoV-2's genome. Based on clues in the genome, they hypothesized the virus is "not a laboratory construct or a purposefully manipulated virus" and pointed to other coronaviruses with similar genetic sequences, including RaTG13. They didn't rule out a lab leak. "We considered the lab leak scenario very closely and looked deeply for any evidence of it in the scientific data," says Anderson.
"In all cases, that data came up supporting a natural origin."
Andersen and his co-authors argue in the March piece that finding related viral sequences in other animal sources would be the best way to reveal where COVID-19 came from. That's because if a similar virus was found, scientists could point to the animal as an "intermediate host," a species that may have facilitated SARS-CoV-2's jump to humans.
Early on in the pandemic, scientists offered a range of species. Snakes famously made it from preprint scientific papers to the mainstream press in January 2020, when information about the coronavirus was still exceedingly scarce. But the analysis was highly flawed and scientists quickly dismissed the idea. If the leak theory was off the table, an intermediate host was likely to exist -- and scientists went hunting.
All the attention shifted to an unusual suspect.
As the only truly scaly mammal on the planet, the slender, ant-eating pangolin is both cute and curious. It's one of the world's most trafficked creatures, hunted for both scales and meat, yet we know little about how it lives.
In February, the pangolin became the first convincing suspect in the hunt for an intermediate host. Previous research had shown that a shipment of sick pangolins, smuggled into China from southeast Asia in March 2019, were suffering from a SARS-like coronavirus. One specific genetic fragment of this virus was similar to that seen in SARS-CoV-2. Such a match, scientists said, made the pangolin a "probable origin" of the pandemic.
Because many of the early COVID-19 cases appeared in Wuhan's now infamous Huanan Seafood Market, where illegal animals were known to be sold, the pangolin seemed to have a plausible link to the outbreak. The market was, at first, implicated as ground zero, but further investigation has shown it likely just allowed the virus to spread effectively during the Wuhan outbreak in December 2019 because it was a gathering point for scores of residents. Records also showed pangolins weren't present at the market.
Yet the enigmatic and unusual creature, so hurt by the illegal wildlife trade, became embroiled in the origin story. "Back in February to at least June last year, people were batshit crazy about pangolins," says Alina Chan, a molecular biologist at the Broad Institute of Harvard and MIT, and a vocal proponent of investigating the lab leak theory.
Three scientific papers in the prestigious journals Nature and Current Biology discussed the pangolin theories in February. It's true the coronaviruses isolated from pangolins show similarities to both RaTG13 and SARS-CoV-2, leading researchers to posit a bat virus and pangolin virus may have swapped genetic material some time ago in a process called recombination, and this may have given rise to the novel coronavirus. This has since been billed as unlikely.
And the pangolin coronavirus data was unusual. Chan and her collaborator Shing Zhan studied the sequences, highlighting a number of inconsistencies between the major studies and questioning missing or unpublished data in a preprint paper posted to bioRxiv. She points to one Nature paper as "dishonest" and says it involves "scientifically unacceptable" practices like publishing samples under different names and the inclusion of deceptive figures. On Nov. 11, Nature added an editors' note to that paper, alerting readers to these concerns. An investigation is ongoing, though the authors have stated these were honest mistakes.
In light of these oddities, and earlier research examining the pangolin coronaviruses, microbiologist Roger Frutos believes the creatures should be "exonerated." Yet, as recently as Jan. 8, the pangolin is still being brokered as a potential starting point in the origins of COVID-19 by Shi Zhengli and other scientists.
Any continued focus on the pangolin, Frutos notes, risks misleading investigations into the origins of the disease. But he also says, even discounting the pangolin's relevance, we may be looking at the emergence of COVID-19 from the wrong angle.
Our understanding of disease emergence revolves around the long-standing "spillover" model.
Spillover occurs when a virus found in wildlife, such as birds or pigs or bats, is able to jump into humans and cause disease.
Bats seem to be excellent drivers of spillover. They are a planet of viruses, hosting many different species that all jostle against the bat's immune system. But bats rarely get sick from the viruses they carry, making them excellent reservoirs. We've seen several high-profile diseases begin in different bat species; Nipah virus and Hendra virus both were isolated in bats. Bats are great coronavirus reservoirs, too.
Frutos believes coronaviruses don't fit into the spillover model perfectly.
Coronaviruses are RNA viruses, a class particularly prone to mutations. The enzyme they use to make copies of their genetic code is error-prone, producing a ton of mutants. "It's the Night of the Living Dead," Frutos says. "You have a huge amount of zombies." These zombie viruses die off quickly, but every now and then, a mistake benefits the virus by giving it an evolutionary advantage, like increasing transmissibility or evasion of the host immune system.
Talking over Zoom, Frutos asks me to put myself in the place of the virus. "For a virus, there are only two kinds of hosts," he says. "A susceptible host and a resistant host." Everything is binary for a virus, he says. Can I infect this cell? Yes or no. Can I replicate? Yes or no.
This thinking led to the development of the "circulation model," an alternative theory to spillover. It hypothesizes that a progenitor to SARS-CoV-2 was likely circulating through a handful of different animals, including humans, before the first cases appeared in Wuhan. Perhaps thousands of zombies had been born in the respiratory tract of a person, before chance and circumstance allowed SARS-CoV-2 to emerge, now-adapted to spread.
It was an "accident," Frutos says.
Sometime in 2019, the accident enabled SARS-CoV-2 to spread through the human population. This probably occurred inside someone who found themselves in Wuhan during the leadup to Lunar New Year. The virus, suddenly adept at infecting humans, now found itself in the middle of one of the planet's biggest annual human migrations. The perfect opportunity to start a global pandemic.
If the origin story is a gigantic jigsaw puzzle, we're still missing some of the most important pieces. The WHO's investigative team, which was set to descend on China in January before the country blocked the team's entry, is tasked with uncovering them.
The task force features 10 researchers, approved by the Chinese government. The WHO terms of reference for the probe say the investigation will be "open-minded, iterative" and will not exclude "any hypothesis that could contribute to evidence generation." It aims to build on work by Chinese researchers and investigators, rather than begin an independent investigation. It makes no specific mention of the WIV or the lab leak theory.
The most contentious scientist on the team is Peter Daszak. As the head of EcoHealth Alliance, a nonprofit that studies spillover events, Daszak has been a collaborator of over 15 years with the WIV's Shi Zhengli, helping fund research and surveilling bat coronaviruses in China to ascertain how the next pandemic might begin.
EcoHealth's surveillance program received millions of dollars in funding from the National Institutes of Health in the US prior to the COVID-19 outbreak. Some of that money has been funneled to the WIV to carry out work on the ground collecting samples from bat caves, mostly in southern China. The collaboration has helped uncover hundreds of bat coronaviruses, some similar to SARS-CoV-2, and also found evidence they might circulate in rural Chinese communities.
The WIV's library of coronaviruses enabled the distant relative, RaTG13, to be discovered quickly after COVID-19 emerged, in turn enabling scientists to home in on mutated genes that made the new virus so potent. Some see these facts as a huge asset to the scientific investigation. There are few more knowledgeable than Daszak regarding Chinese laboratories and the emergence of disease in the region.
But Daszak's close relationship with the WIV is also seen by many as a conflict of interest when it comes to the WHO's investigation. Rowan Jacobsen, a freelance journalist who has investigated the leak theory, said it was like "having Donald Trump run the investigation of Russian hacking of the 2016 election." A WHO spokesperson said "all experts participating in WHO expert groups are signing declarations of interest."
Daszak did not respond to repeated requests for comment. When BBC journalist John Sudworth asked him about this perceived conflict, he responded, "We file our papers, it's all there for everyone to see."
The lab leak theory initially hurt EcoHealth Alliance's operations in China. Shortly after the pandemic began, the NIH cut funding to the organization. Then, in August, a $7.5 million grant was awarded to EcoHealth to continue this work. If a virus did escape -- accidentally -- from the WIV, there's a lot on the line for the firm.
"A lab leak situation could directly threaten all of that," says Sainath Suryanarayanan, a staff scientist at investigative nonprofit US Right To Know looking into the origin story. This should not be taken as evidence of a vast conspiracy spearheaded by Daszak and the Chinese to cover up a lab leak. It merely highlights the conflicts of interest presented by Daszak's inclusion.
Under these circumstances, can the investigation hope to find any evidence of a leak? "I have zero confidence left in the WHO team," Chan says.
What makes discussing the lab leak theory so uncomfortable -- and potentially dangerous -- is how often it's aligned with more extreme, often sinophobic or nationalistic conspiracy theories around the emergence of SARS-CoV-2.
An intentional release of an engineered bioweapon and an accidental leak are two markedly different events, but they have become intertwined, driven by a misplaced motivation to shift the blame to China, as well as politicking, a polarized online discourse and a lack of certainty that has undercut communication around the pandemic since its very beginning.
This lack of certainty has allowed conspiracy theories to fester, particularly on social media, where misinformation and a skewering of facts are prevalent. "It's not necessarily a conspiracy to consider a lab leak scenario," says Andersen, the virologist from Scripps, "but most theories about lab leaks are indeed conspiracy theories."
For instance, the bioweapon theory, comprehensively refuted time and again, is intricately tied to a laboratory-based origin. But it's based on bogus facts requiring significant leaps of logic. Other theories, suggesting the rollout of 5G was somehow linked to the emergence of COVID-19 or that Bill Gates is using the virus to usher in a new world order, also have no foundation in truth.
These are conspiracies. But some arguments against an accidental leak have unfairly mixed the conspiracies with legitimate lines of inquiry. An accidental leak should be given equal weight, when there's no clear evidence against it. Yet the mess of adjacent conspiracies has seen most scientists shy away from the topic altogether.
"I think it is plausible that either SARS-CoV-2 emerged 'naturally' from some sort of interaction between humans and animals, or that it was an accidental release from a lab," says Jesse Bloom, an evolutionary biologist at Seattle's Fred Hutchinson Cancer Research Center. "Beyond that, I think we quickly get into the realm of speculation, which is not something that I want to do."
Some scientists dragged into the origins debate told me they preferred to focus on other aspects of the pandemic. "Why speculate fruitlessly on conspiracies when there are real problems to solve and real people to help, now," says Magdalena Plebanski, professor of immunology at RMIT University in Melbourne, Australia. Others alluded to the toll it was taking on their mental health.
It's no secret why. Those who do speak out for or against the lab leak theory find themselves in the middle of fiery clashes and heated exchanges, some playing out in the public eye via Twitter; others more clandestine, via inboxes and private messages. The constant questioning is wearing down researchers. For scientists like Andersen, countering misinformation around the origin story has been "extremely distracting and time consuming." One prominent virologist, Angie Rasmussen of Georgetown University, wrote in Nature Medicine that she has experienced threats of violence and sexual assault for debunking misinformation.
Chan has gone from relative obscurity to uneasy infamy during the pandemic, trying to piece the origin story together from a laptop, tweeting out inconsistencies in scientific literature and asking questions as new information comes to light. For the lab leak believers, she's been a beacon. But she admits it's often been difficult to ask "moderate questions" in places like Twitter, where polarization immediately pushes those inquiries to the extreme.
She's had trouble sleeping, afraid speaking out could harm her friends, colleagues or family. "There's a lot of lows," she says.
Pandemics are inevitable, unavoidable accidents, but they are rare. They require a unique combination of low-probability events to align. Until December 2019, we had been extremely lucky.
We can't rely on luck to prevent the emergence of the next virus. Tracing the origin story back to the very beginning provides a learning opportunity. We will not be able to prevent the next accident from occurring. "We cannot predict the next epidemic," says Frutos, the biologist at CIRAD.
That makes the hunt for COVID-19's origins all the more urgent. Whether SARS-CoV-2 jumped into humans from a bat or escaped from a laboratory is critical to managing the next emerging disease before it can become a pandemic.
To date, countless papers and endless hours have investigated the natural origins theory. History shows this very likely could be the source of SARS-CoV-2. The lab leak theory became so entangled in conspiracy and geopolitics that it was almost immediately dismissed. A serious, reputable investigation is required.
The WHO mission in Wuhan is not prepared to do so. It will, a spokesperson says, examine hospital records and map activities and items traded at Huanan and other seafood markets. But even if the investigation was giving serious thought to alternative hypotheses like an accidental leak, so much time has elapsed since COVID-19's emergence that we may have missed our window. "I'm afraid it's too late for the lab leak theory to be seriously investigated," says Suryanarayanan, from US Right To Know. "That said, I don't think people should give up on that."
Chan certainly won't. "There are things that people can go hunting for now that are not inside of China, that don't need Chinese permission," she says. She is dogged in her pursuit of new clues into COVID-19's beginning, pounding away on a keyboard, digging through reports and stitching together any thread of evidence she can find. She thinks there is good reason to be optimistic.
There are still clues to be found.
Want to get in touch about COVID-19's origin story? Email the author.
Originally published Jan. 19.
Correction: A previous version of this article said there were 9 cases of SARS during the 2004 "leak" -- there were 11. Thanks to Gilles Demaneuf for the tip. The shipment of smuggled pangolins is from March 2019, not 2017 and 2018.