November 2023


This little theory went to market

By Elizabeth Finkel
Close up of person in PPE carrying presumed infected material in a plastic bag at Huanan Seafood Market, January 27, 2020

Huanan Seafood Market, January 27, 2020. © An Yuan/China News Service via Getty Images

Using the scientific method to debunk the persistent claim that Covid-19 originated in Wuhan as a lab leak

Reporting on the origins of the Covid pandemic has been a challenge for journalists. From the beginning there have been two narratives. One was that the virus, SARS-CoV-2 (SARS-2), jumped from an infected animal into humans at the Huanan Seafood Market, in the city of Wuhan, China. The other was that the virus was genetically engineered at one of the two campuses of the Wuhan Institute of Virology, which lie 12 and 27 kilometres to the south of the market across the Yangtze River. After accidentally infecting a lab worker, it spread into Wuhan and ultimately the world.

In short: natural origin versus a lab leak.

Some journalists became proponents of the lab leak and were taunted as “conspiracy theorists”; others followed the mainstream scientific view and were taunted as “stenographers of science”.

My inclination was natural origins, given that the first SARS outbreak had arisen this way, as had Ebola, Middle East respiratory syndrome (MERS), HIV, yearly bird flu outbreaks and others. But the accusation of “playing stenographer” stung. It forced me to check my process. I do not blindly trust scientists. My lodestone is the scientific method itself.

It works like this. You form a hypothesis. You gather data to test it. That doesn’t mean only trying to validate it, it also means trying to see if you can disprove it. Most of the time, science can’t prove things; it advances based on the “weight of evidence”.

What I want to do here is show how I weigh the evidence based on the best two sources I can muster: the peer-reviewed scientific literature and, in a novel experience for me, the United States intelligence agencies that reported their findings to the Biden administration in June.

The bottom line: the evidence weighs heavily on the scale towards natural origins, leaving the lab-leak side a lightweight proposition.

Before I take you through the evidence, it’s instructive to revisit the first SARS outbreak beginning at the end of 2002. It claimed a few lives in Europe and the Americas, taking its major toll in Hong Kong and China: at least 770 deaths. It seems small fry by comparison with SARS-2 that claimed at least six million. But SARS-1 was a wake-up call. Till then, coronaviruses seemed relatively harmless, agents of maladies such as the common cold. After SARS -1, the international research community mobilised to try and stop such a thing from happening again.

The key was to figure out where SARS-1 had come from. The first part of the answer was easy: palm civets, weasel-like animals, sold at street markets in Guangdong. Here, live animals, often trapped in the wild but nominally produced by the wildlife farming industry, were butchered and consumed at nearby eateries. But civets could not provide the whole story. No significant wild or farmed population outside of the markets seemed to carry the virus, so its evolutionary source likely lay elsewhere. For virologist Linfa Wang, formerly at CSIRO, now at Duke-NUS Medical School in Singapore, it was most likely a bat. Unlike other mammals, bats show remarkable tolerance to coronaviruses, allowing them to proliferate without themselves sickening. Collaborating with Zhengli Shi’s lab at the Wuhan Institute of Virology (WIV), researchers searched bat caves for SARS-like viruses. Some of the research was funded by the not- for-profit US-based EcoHealth Alliance headed by Peter Daszak.

In 2013, the researchers found the source in bat caves in Yunnan province in China’s south-west: coronaviruses that were 95 per cent similar to SARS-1. They also found something alarming. The viruses were evolving fast, producing weird new combinations. Tests at WIV and at Ralph Baric’s lab at the University of North Carolina, showed these novel combinations were able to infect human cells.

Warnings were issued. The EcoHealth Alliance produced pamphlets to educate villagers on how to live safely around bats. And in international publications from 2013 to 2017, the international team – Shi in China, Baric and Daszak in the United States, and Wang in Singapore – warned of the clear and present danger.

In December 2019, SARS-2 emerged in Wuhan.

Over half of the initial cases were people who worked at the Huanan Seafood Market or had been in contact with someone who had. On January 1, 2020, the market was shut and disinfected. By February 24, a national law was put in place outlawing the eating of wild animals and the wildlife trade that supplied them. It seems China was quite clear about the origins of SARS-2: it had jumped ship from a wild animal at the Huanan Seafood Market.

Yet by February 2020, the world was already entertaining another hypothesis: the lab leak.

Ironically, the two researchers who first raised the possibility now find themselves accused of trying to cover it up.

Eddie Holmes, at the University of Sydney, and Kristian Andersen, at Scripps Research in San Diego, are leading lights in the field of virus evolution. In August, the Royal Society in the United Kingdom recognised Holmes’s contribution with the Croonian Medal, the same prize Howard Florey received for developing penicillin. These two boffins spend their waking hours poring over the family trees of viruses.

In mid January 2020, the pair took their first look at the 30,000-letter genetic code of the SARS-2 virus – and got a shock. There were two strange signatures not seen in SARS-1. Holmes was alarmed by a 20-letter string of code lying within the coronavirus’s homing device, known as the receptor binding domain. Like SARS-1, it targeted a receptor on human lung cells, but with even better precision. Andersen was alarmed by a 12-letter string of code that was not present in SARS-1 at all. Known as a furin cleavage site, it helped unleash the homing device. Both new signatures might explain the super-spreading power of this new virus.

Could the signatures have arisen from genetic engineering experiments at the WIV? The Wuhan institute was a world-renowned sentinel for monitoring bat viruses. Its freezers contained thousands of bat viruses and it conducted experiments to test their function. Could such a test have gone awry?

Their concerns quickly reached Anthony Fauci, then head of the US National Institute of Allergy and Infectious Diseases (NIAID) and stoic adviser to President Donald Trump. Fauci recommended two things: that a conference call be conducted including some of the leading experts on virus evolution and public health, and that they contact their respective intelligence agencies to share their concerns. In Australia, Holmes contacted Nick Warner, then head of Australia’s Office of National Intelligence.

The experts got together on February 1. They brought new data mined from virus databases. During the course of that meeting, which included a mostly silent Fauci, and over the next few days, Holmes and Andersen changed their minds – they concluded that SARS-2 was not likely to be the result of a lab leak. Exactly what scientists are supposed to do when faced with new data.

But that change of mind gave birth to a conspiracy theory. Journalists such as Australia’s Sharri Markson join the dots as follows: Fauci’s agency NIAID had funded work at the WIV. So Fauci was trying to cover up the evidence of the lab leak to prevent reputational damage. Andersen was the recipient of part of a US$9 million grant from NIAID, so he was motivated to collude with Fauci. And Holmes? Markson has repeatedly accused him of being part of the cover-up, but it’s not clear what motive he’s supposed to have had.

None of this constitutes evidence. They are ad hominem attacks. But for the sake of the exercise, let’s test how plausible these assertions are. Fauci claims he had no idea of the NIAID funding to the Wuhan institute, pointing out that his department funded projects to the tune of US$6 billion per year. The institute received US$120,000 per year, subcontracted via the EcoHealth Alliance, an amount hardly likely to come to his notice.

Second, the claim that Andersen was colluding with Fauci to protect his grant? Indeed, a $9 million grant would have been more noticeable, but NIAID grants are not awarded by fiat of Fauci. There is an independent peer-reviewed process. Andersen’s grant had scored winning marks a month before the February 1 meeting.

And, while we’re dispensing with the lightweight stuff, let’s include another bit of “evidence” typically offered in favour of a lab leak: the claim that the first Covid-19 patient in Wuhan was a researcher at the WIV. US intelligence agencies looked for some substantiation of that and found none.

What was the evidence that persuaded Holmes and Andersen to change their minds in February 2020?

After searching the databases, it turned out that the two alarming signatures in the genetic code of SARS-2 were not unusual for coronaviruses. Regarding the furin cleavage site, MERS had one, and so did the common cold coronaviruses. The receptor binding domain in the virus homing device was not new either. It turned up in a coronavirus that had been isolated from illegally traded pangolins. If nature was cooking up these code changes in coronaviruses, there was no longer any compelling reason to propose a man-made origin.

A second line of evidence against a laboratory-made virus came from considering how plausible it would have been to create it. For starters, if genetic tinkering at WIV had produced SARS-2, then what starter virus was used as its chassis? While the institute had thousands of bat viruses in its freezers, most were not intact viruses but fragments extracted from bat poo. Only three viruses were intact enough to be grown in animal cells. One was a SARS-1 ancestor, called WIV-1. But it could not have served as the chassis for SARS-2: it was only 79 per cent similar across the 30,000 letters of code.

Zhengli Shi, frantic that her lab might be responsible for an accidental leak, examined all the genetic codes that had been read for the bat coronavirus in her freezers. One gave her pause. Isolated in 2013, its genetic code was a 96 per cent match to SARS-2. But the virus, named RaTG13, was one of those Shi had never been able to grow in animal cells. If she had, we’d have heard about it. Her stellar CV included regular publications on newly discovered viruses in Nature and Science.

Armed with that evidence, Holmes and Andersen changed their thinking over a few days in February. The balance of probabilities showed that this new virus was not made in the laboratory of man, but in nature.

In the three and a half years since, the evidence has mounted in favour of that assessment. We’ll get to it in a moment.

But first, we need to examine a more substantive piece of evidence for the lab-leak hypothesis.

In September 2021, a group of lab-leak sleuths unearthed a failed grant proposal titled “DEFUSE”. Submitted by the EcoHealth Alliance in 2018 to the US Defense Advanced Research Projects Agency (DARPA), its ultimate goal was to “defuse” the risk of virus spillovers from bats. Part of that work involved taking bat viruses and engineering different furin cleavage sites into them to test their effects. The work was to be a collaboration between the WIV, which would supply the bat viruses, and Ralph Baric’s group at the University of North Carolina, who would do the genetic engineering. DARPA rejected the grant proposal. Baric says the work did not go ahead.

Nevertheless, reports of the DEFUSE grant hit like a bombshell.

The lab-leakers saw it as the smoking gun that explained how SARS-2 acquired its extraordinary furin cleavage site. Even if DARPA failed to fund the project, the plans were in place. It might have gone ahead.

But evidence that it happened? Two years on there is none.

Some hypothesise that the DEFUSE work took place at the WIV and the virus leaked, because work there was undertaken at a moderate biosecurity level, rather than the higher biosecurity level that would have been used in North Carolina.

But that argument still fails the test of what starter virus was used. Even if researchers had been able to find a way to grow the close relative of SARS-2 – RaTG13 – there are over 1000 letters of genetic code that are different. Experts such as Stuart Turville at the University of New South Wales, who engineers new viruses for gene therapy, says the odds of being able to artificially introduce such huge chunks of code are extremely low. Wang puts it more bluntly. Until nature had showed us what it took, he tells me, “no scientist, smart or rich, would be able to make SARS-CoV-2”.

And even if you believe, as the lab-leak champions do, that Shi is covering up – she has, for instance, declined to surrender her WIV lab books – that is not true for Baric. The intelligence agencies would have had free access to his lab. Their finding was that no chassis existed.

Hence, an evidence-based approach leads to dismissal of the DEFUSE grant argument.

Finally, let’s get to the scientific evidence that has emerged over the past three years that weighs heavily in favour of a natural origin for SARS-2.

First, the hunt for the bat ancestors is proceeding much like the hunt for the SARS-1 ancestor did. Researchers then never found a perfect match – the closest, YN2020E, was 96 per cent similar. Rather, they found several relatives, each of which carried a different part of the SARS-1 virus. The most likely explanation is the coronaviruses swapped bits to produce the ancestor of SARS-1.

So far, the hunt for SARS-2 ancestors has likewise delivered several family members. French researchers found one group of viruses in bat caves in Laos. Dubbed BANAL, they showed a 96.8 per cent match overall, and almost a 100 per cent match over the virus homing device, the receptor binding domain. Another bat coronavirus found in Yunnan province, RmYN02, showed a lower 93 per cent match overall, but a 97 per cent match across a SARS-2 gene called “1ab”. Examining the genetic codes of these relatives, Jonathan Pekar, a bioinformatics expert at the University of California San Diego, calculates the SARS-2 ancestor may have circulated as little as three years prior to the outbreak.

None of the bat coronaviruses so far appear to show a match to the furin cleavage site. Pekar believes the most likely scenario is that this bit of code evolved in an intermediate animal host, perhaps a palm civet or raccoon dog – animals susceptible to SARS-2 infection and known to be sold at the Huanan Seafood Market.

We may never discover the original host. Whatever infection the species carried has likely been swamped by the SARS-2 that jumped from humans to animals – it has become rife in populations from American white-tailed deer to Dutch mink.

But the major animal host for SARS-2 – humans – reveals something interesting. As SARS-2 has evolved from the original Wuhan strain to Delta, Omicron and now Pirola variants, the furin cleavage site keeps changing. For Gary Whittaker, a Cornell University virologist who tracks the evolution of this chunk of code, it’s evidence that the furin cleavage site is “the design of nature, not man”.

A completely different line of evidence comes from evolutionary biologist Michael Worobey at the University of Arizona. Worobey has a reputation for trying to “disprove” theories. Two decades ago, he travelled to the Democratic Republic of Congo to disprove the dominant theory that HIV crossed over from wild chimps – the alternative being that it crossed over from a contaminated polio vaccine. His findings proved the dominant theory correct.

With the outbreak of SARS-2, Worobey was suspicious about reports linking the earliest cases to the Huanan Seafood Market. Perhaps an unseen factor was skewing the results? For instance, market workers might have been the first to be tested. To disprove the market origin hypothesis, he mapped the home addresses of the first 150 cases onto a map of Wuhan, a sprawling megacity of 11 million. The addresses clustered around one spot: the seafood market. “Long story short, they all lived absurdly close to and centred on the wildlife market; the highest probability of drawing a case is within a bubble a 100 metres from the market,” Worobey told me. His paper providing these findings was published in Science in 2022.

But was the market simply the site of a super-spreading event? If so, then it should have spread a single strain of the virus, but another paper in Science in 2022, by Jonathan Pekar, showed the market harboured not just one strain, but two. That’s far more consistent with what is known to happen with animal viruses, which make hopeful leaps to humans all the time.

China has been completely unhelpful about resolving the origins debate. Its authorities declined to hand over researchers’ lab books, but they also declined to share crucial data from the Huanan market. For instance, we now know that swabs taken from market prior to the shutdown contained DNA from several wild animal species susceptible to SARS-2, including raccoon dogs and civets. Many of the swabs also contained SARS-2 – implicating one of these species as the original carrier. Yet it took more than three years for this evidence to be made publicly available.

Why has China been so unhelpful? Competing interests, says Peter Li, a wildlife policy scholar based at the University of Houston-Downtown. The wildlife trade – once a US$74 billion industry – is fighting for survival. Li says the lab-leak theory has helped stage a “spirited comeback” because it exonerates the industry from blame.

Li traces the beginning of China’s stonewalling to May 2020, when Trump’s trade adviser Peter Navarro accused China of seeding Covid-19 by sending “hundreds of thousands of Chinese on aircraft to Milan, New York and around the world”.

Toxic politics from both sides have damaged science. The vibrant collaborations between Chinese and Western scientists that constituted an early warning system for emerging viruses have been severed. And in their own countries, Holmes, Andersen and colleagues are sapped of energy as they fend off media smear campaigns and the harassment and death threats that follow. We are arguably in a far worse state of readiness for the next inevitable pandemic.

The media bear much of the responsibility. As the scientific evidence mounts in support of a natural origin, public opinion veers in the opposite direction. Polls show that some two thirds of Americans believe the lab-leak hypothesis is correct. Journalists need to examine their process. Rather than wedding themselves to a particular narrative, they can do no better than taking a page from the scientific method.

Elizabeth Finkel

Melbourne-based Elizabeth Finkel is a former biochemist who switched to telling the stories of other scientists. She is the former editor of Cosmos magazine.  

From the front page

Members of the Kanakanvu tribe perform at a Saraya harvest festival, Donghua Village, Taiwan.

Who is Taiwanese?

Taiwan’s minority indigenous peoples are being used to refute mainland China’s claims on the island – but what does that mean for their recognition, land rights and identity?

Image representing a film still of abstract colours

Tacita Dean and the poetics of film editing

The MCA’s survey of the British-born artist’s work reveals both the luminosity of analogue film and its precariousness

Image of David McBride

David McBride’s guilty plea and the need for whistleblower reform

The former army lawyer had no choice but to plead guilty, which goes to show how desperately we need better whistleblower protections

Illustration by Jeff Fisher

Mars attracts

Reviving the Viking mission’s experiments may yet find life as we know it on Mars, but the best outcome would be something truly alien

In This Issue

Illustration by Jeff Fisher

There are no words

Finding solace and surprise around a family Scrabble board with a grandmother in decline

A public-housing brick three-storey building in Ascot Vale

A house provided: Preserving public housing

The architectural practice proving that refurbishing public housing can be less expensive and disruptive than demolition for new projects

Michael Fassbender in ’The Killer’, sitting in a room cross-legged on a mat, wearing black gloves

Into the streaming void: ‘The Killer’ and ‘They Cloned Tyrone’

David Fincher’s stylish pulp and Juel Taylor’s SF-adjacent satire are the latest riches to be taken for granted in the ever-ready, abundant world of Netflix

Empty seats with No campaign placards on them in an event venue in Melbourne, September 15, 2023,

True colours

What the outcome of the Voice referendum suggests about the future of reconciliation, and what it says about the national character

More in The Monthly Essays

Close-up photograph of Anne Summers, 2017

How to change a bad law

The campaign to repair the single parenting payment was a model of how research and advocacy can push government to face the cruel effects of a policy and change course

Photo of Animal Justice Party MP Georgie Purcell outside Parlament House with rescued greyhound Graham

Dog day afternoon

Animal welfare concerns have long plagued the greyhound racing industry, but in Victoria a campaign from covert investigators now has a parliamentarian leading the fight

Kakadu National Park

Park of the covenant

With the Ranger Uranium Mine now closed, Kakadu’s traditional owners want the government to make good on the original promise of a national park in their care

Ningaloo, Western Australia

Writing in nature

The laureate of the Western Australian surf break sees a task ahead for our essayists, our novelists and our poets alike in the fight against climate catastrophe

Online latest

Image representing a film still of abstract colours

Tacita Dean and the poetics of film editing

The MCA’s survey of the British-born artist’s work reveals both the luminosity of analogue film and its precariousness

Image of David McBride

David McBride’s guilty plea and the need for whistleblower reform

The former army lawyer had no choice but to plead guilty, which goes to show how desperately we need better whistleblower protections

Installation view of the Kandinsky exhibition at the Art Gallery of New South Wales, showing three framed abstract paintings hanging on a wall

Kandinsky at AGNSW

The exhibition of the Russian painter’s work at the Art Gallery of NSW provides a fascinating view of 20th-century art’s leap from representation to abstraction

Image of Margret RoadKnight playing guitar and singing.

The unsung career of Margret RoadKnight

Little-known outside the Melbourne folk scene for decades, singer Margret RoadKnight’s 60 years of music-making is celebrated in a new compilation