麻豆传媒

Before heading off to China as leader of a World Health Organization (WHO) fact-finding mission into the origins of SARS-CoV-2, Peter Ben Embarek recorded an outlining the state of knowledge at the time, January 2021.

鈥淲e know that the first human cases that were detected were detected in Wuhan in December 2019,鈥� he said. 鈥淲e also know that this virus belongs to a group of viruses that have their original niche in bat populations. In between these two points, we don鈥檛 know much.鈥�

Almost six months on, we still don鈥檛 know much. Arguably, we actually know less, with the two 鈥渒nowns鈥� now being called into question. Even though Embarek鈥檚 investigation concluded that one of the possible origins of SARS-CoV-2 鈥� accidental release from a laboratory 鈥� was 鈥渆xtremely unlikely鈥�, that dreadful possibility still hasn鈥檛 been ruled out. If anything, the case for a lab leak has grown stronger.

Last week, The Wall Street Journal ran an claiming that US intelligence has evidence of several employees of the Wuhan Institute of Virology, which has long carried out research on potentially dangerous bat coronaviruses, being hospitalised with a respiratory illness very similar to covid-19 in November 2019. US President Joe Biden subsequently ordered the US intelligence community to pursue a definitive conclusion on whether the virus spilled naturally from a wildlife reservoir, or unnaturally from a lab.

The origin of the virus remains one of the biggest, most important and most contentious unknowns of the pandemic. 鈥淎bsolutely we need to know where it came from,鈥� says , an evolutionary virologist at the University of Glasgow, UK. 鈥淲e have to be worried that that could happen again.鈥�

So what is the evidence for and against a laboratory leak? And what pieces of additional scientific evidence are required to adjudicate on the matter?

Consensus on a natural origin

For now, there is a near-consensus that SARS-CoV-2 had a natural origin in a wild animal, says microbiologist Rossana Segreto at the University of Innsbruck in Austria.

That consensus is the one strongly favoured by Embarek鈥檚 WHO investigation. At a press conference at the end of the mission in Wuhan on 9 February, he said that the virus seems to have originated in bats as originally thought. The WHO鈥檚 , published on 28 February, reiterated the bat origin hypothesis.

However, on 4 March, a group of scientists published an in The New York Times calling for an independent investigation on the grounds that the WHO 鈥渄id not have the mandate, the independence, or the necessary accesses to carry out a full and unrestricted investigation into all the relevant SARS-CoV-2 origin hypotheses鈥� 鈥� including the lab leak. A few weeks later, the governments of 14 countries including the US, UK and Australia that the WHO investigation 鈥渓acked access to complete, original data and samples鈥�.

Earlier this month, the journal Science from a group of 18 distinguished scientists entitled 鈥淚nvestigate the origins of COVID-19鈥�. It argued that theories of accidental release from a lab and so-called zoonotic spillover (where an infectious disease jumps from an animal to a human) 鈥渂oth remain viable鈥�.

One of the signatories is David Relman at Stanford University in California, who has if only to debunk it. 鈥淭here鈥檚 still a lot of scientists, in my view, who are a bit locked into the assumption that this has and can only have a natural origin,鈥� he says. 鈥淚鈥檓 not quite sure why.鈥�

Fuel for doubts

A lot of the doubts are fuelled by dissatisfaction with the WHO investigation and suspicion of ulterior Chinese motives. The WHO team had a 鈥渞eally difficult job鈥�, says Robertson, because 鈥渢he Communist party of China want to project it out of China鈥�.

But there are also scientific reasons to question the consensus. 鈥淪everal characteristics of SARS-CoV-2 taken together are not easily explained by a natural zoonotic origin hypothesis,鈥� writes Segreto in .

The lab-leak hypothesis usually points the finger at the Wuhan Institute of Virology, which is close to the Huanan Seafood Market,where the first major cluster of infections occurred. The institute has a long history of collecting and analysing bat coronaviruses. The leak scenario usually involves researchers tinkering around with a virus to investigate its properties, perhaps in 鈥済ain of function鈥� experiments in which pathogens are modified to be more harmful in a bid to understand them better. This modified virus then somehow slipped through the lab鈥檚 biosafety net, which has been for being full of holes.

Robertson points out that there is no documented evidence of such experiments taking place. The WHO team granted access to the institute found none. The Wuhan Institute of Virology has reported working with a , which is the closest-known relative of SARS-CoV-2 with a genome sequence similarity of 96.2 per cent. But even this is genetically quite distant from SARS-CoV-2 and RaTG13 clearly isn鈥檛 its immediate progenitor, says Robertson. 鈥淭hey weren鈥檛 working on the right viruses,鈥� he says.

That, of course, doesn鈥檛 rule out undocumented experiments. There are good reasons to believe that the institute hasn鈥檛 always been entirely transparent, says Relman. For example, in November last year, it published a to a on RaTG13 revealing that sampling missions to a copper mine in Yunnan Province where that virus was discovered also yielded eight other previously unknown SARS-like coronaviruses. The addendum didn鈥檛 give any further details of these viruses.

Unknown virus

Intriguingly, the Wuhan Institute of Virology was alerted to the Yunnan site in 2012 when four miners fell ill with a mysterious respiratory illness after going into the mine to clean up bat guano. One of the men died of his illness. The institute subsequently tested samples from the men and confirmed that they weren鈥檛 infected with SARS-CoV-2, but hasn鈥檛 determined what caused the illness beyond suggesting that it was 鈥渁n unknown virus鈥�.

The original omission, and subsequent admission, of this information hasn鈥檛 been adequately explained, says Relman. 麻豆传媒 emailed Zheng-Li Shi, head of bat coronavirus research at the Wuhan Institute of Virology and lead author of the addendum, for comment but she didn鈥檛 reply.

But to go from there to positing secretive experiments that ended horribly is to enter the realms of speculation, says Robertson. 鈥淚t loses all meaning at that point because it鈥檚 not about facts any more. Unless you have some solid evidence that they were working on viruses very closely related to the one that 鈥渆scaped鈥�, then that鈥檚 where it becomes a conspiracy theory.鈥�

However, proponents of the lab-leak hypothesis can point to some arcane details of the virus鈥檚 molecular biology. None of these is a smoking gun on its own, but taken together, they challenge the natural origin hypothesis, argues Segreto.

For example, the virus has a 鈥渇urin cleavage site鈥�, a part of the spike protein that helps it to break into host cells. Many coronaviruses have this tool, but SARS-CoV-2 is the only member of the sub-genus Sarbecovirus to have one.

Another region of the spike protein, the 鈥渞eceptor binding motif鈥�, appears to be oddly adapted to latch on to human cells. This adaptation was also observed in the original SARS virus, SARS-CoV-1, which caused outbreaks in 2003, but only long after it had jumped to humans. The Wuhan strain of SARS-CoV-2 had it from the get-go, as if it were 鈥減re-adapted鈥� to humans.

Common adaptations

These and are theoretically consistent with a virus that has been manipulated in the laboratory, says Segreto, possible by a process called 鈥渟erial passage鈥�, whereby the virus is adapted to humans by making it infect cultured cells, selecting variants that succeed, and repeating the process.

Not so fast, says Robertson. 鈥淭he 鈥榠t doesn鈥檛 look like it鈥檚 natural鈥� claim is preposterous, because all of those features, the furin cleavage site and the receptor binding motif, they鈥檙e all very typical, you can find them in natural viruses鈥�. An almost-identical section of the furin cleavage site, for example, was recently discovered in a bat Sarbecovirus from Thailand, says Robertson.

Another recent by researchers at ShanghaiTech University in China reports that furin cleavage sites are common across the coronavirus family and appear to have evolved independently multiple times in different lineages. This supports the natural origin hypothesis, say the authors.

The superficial appearance of unnaturalness arises, says Robertson, because of a phenomenon called recombination. The enzyme that copies the viral genome is highly promiscuous, and in a mammal cell co-infected with two coronaviruses, it can stitch together bits of both viral genomes in novel combinations. This can cause incongruous molecular features to suddenly appear in a virus lineage as if by magic, or design.

Factor in recombination and it is possible to construct a perfectly natural evolutionary tree of sarbecoviruses including SARS-CoV-2. 鈥淲hat鈥檚 very clear to anybody that鈥檚 worked in this field is that SARS-CoV-2 is really just another sister lineage to that first SARS virus that first emerged in 2002,鈥� Robertson says.

As for 鈥減re-adaptation鈥�, there is nothing to see, says Robertson. The virus merely evolved to be a generalist, enabling it to extend its natural range beyond bats and into other mammals, which just so happens to include humans. The virus readily infects many other species including mink, pangolins and cats.

No smoking gun

Robertson admits that the smoking gun of the natural origin hypothesis is also absent. That would be a naturally occurring virus that is genetically close enough to SARS-CoV-2 to plausibly be its direct ancestor. 鈥淚t remains most likely that the immediate ancestor to SARS-CoV-2 exists in the wild and is still to be found,鈥� says Jonathan Stoye at the Francis Crick Institute in London.

But Robertson points out that searching for such a progenitor is like looking for a needle in a bat cave. Bats carrying SARS-like coronaviruses live right across China and into South-East Asia, and current levels of sampling aren鈥檛 adequate.

He is also at pains to point out that he and his colleagues will follow the science where it leads. 鈥淚f there was a piece of good evidence [for the lab leak hypothesis] that came out tomorrow, we would pivot on that very quickly.鈥�

All things considered, both hypotheses have to be left on the table for now. Work is ongoing to reject one or the other, not least by Embarek鈥檚 WHO team, which is far from finished with its investigations. Biden has given his intelligence agencies 90 days to report back. But bear in mind that it took a decade to discover the origins of SARS-CoV-1, which was unimpeded by geopolitical intrigue. Don鈥檛 hold your breath.

Sign up to our free Health Check newsletter for a round-up of all the health and fitness news you need to know, every Saturday