The scientists who assured the world that the Covid-19 virus could not have been engineered in a laboratory based their pivotal decision on a single piece of flawed evidence. Their discussion of the scientific facts was interspersed with frequent speculation about the public impact of their findings. Theirs was no openminded search for the truth; one scientist expressed his determination from the start to disprove the possibility of a lab leak. In the rush to publish their predetermined conclusion, they ignored a critical viral feature that points to manipulation.

These departures from customary scientific procedure are evident from a new batch of emails released to Jimmy Tobias, a freelance writer, after litigation to block the National Institutes of Health (NIH) from making extensive redactions. The emails were exchanged principally between Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), a branch of the NIH, and Jeremy Farrar, director of the Wellcome Trust, a large medical research foundation in London. Other participants were Francis Collins, then director of the NIH, Patrick Vallance, Chief Scientist to the British government, and two groups of virologists, one largely U.S.-based and the other from Europe.

The new emails date from the first nine days of February 2020. They record the participants’ discussions of how to frame a report discrediting the idea that the Covid virus, SARS-CoV2, could have escaped from a lab. Just such a proposal had been outlined by the American virologists in a January 31 email to Fauci. Strangely, two of the American group soon began to lead the charge against their own proposal. Their efforts led to preparation of an article, overseen by Farrar, Fauci, and Collins, in which the American virologists reversed themselves entirely, declaring it impossible that the virus could have been engineered. The article, known as the Proximal Origins paper, was posted in near-final form as soon as February 17. It was published in the journal Nature Medicine on March 17, 2020, and proved widely influential in promoting the view that SARS-CoV2 had evolved naturally.

The emails begin with vigorous statements by the American virologists explaining their initial view that the virus had indeed been concocted in a laboratory. Mike Farzan of Scripps said that he was “bothered by the furin site” and had “a hard time explaining that as an event outside the lab.” The SARS-CoV2 genome, some 30,000 nucleotide units in length, contains a 12-nucleotide insert, known as a furin cleavage site, which greatly enhances its infectivity. Closely related viruses frequently exchange genetic material, so it would be easy to see SARS-CoV2’s furin cleavage site as having a natural origin if any other viruses in its group possessed one. But none does. Hence Farzan’s perplexity and his inference that the furin site must have been engineered into the virus.

The same point worried Bob Garry, a virologist at Tulane University. “I really can’t think of a plausible natural scenario . . . where you insert exactly 4 amino acids 12 nucleotides . . . I just can’t figure out how this gets accomplished in nature,” he wrote.

But the European virologists present on a February 1 teleconference convened by Farrar were dismissive of the lab-leak hypothesis, as if they had been invited to rebut the foolish presumptions of their American colleagues. They included Christian Drosten of Germany and Ron Fouchier, a Dutch virologist who raised alarms in 2011, when he created an air-transmissible strain of a killer flu virus, funded by the NIAID. Both virologists argued strongly that a natural origin of the Covid virus was more likely. Collins in one email complained to Farrar that “the arguments from Ron Fouchier and Christian Drosten are presented with more forcefulness than necessary,” though he was nonetheless coming around to their view.

The most surprising interventions in the discussion came from the other two members of the group, besides Farzan and Garry, who had warned Fauci of the virus’s artificial nature—Kristian Andersen of Scripps and Edward Holmes of the University of Sydney. Almost overnight, their view shifted from concluding the virus was manufactured to arguing that this could not possibly be the case. On February 8, Andersen told the others that “Our main work over the last couple of weeks has been focused on trying to disprove any type of lab theory.” So much for any open-minded inquiry into the virus’s origins.

Andersen went on to complain that so far, the scientific evidence wasn’t conclusive enough to say whether the virus acquired its distinctive features through natural selection—whether in humans or an intermediary animal host—or by serial passage, but that he was “very hopeful that the viruses from pangolins will help provide the missing pieces.” Serial passage refers to a laboratory manipulation in which a virus is forced to adapt to new conditions by continually passing it from one medium, whether a cell culture or lab animal, to another.

Holmes, too, argued against the lab-leak thesis, so much so that Collins said the possibility of serial passage should still be kept in mind. “I’d be interested in the proposal of accidental lab passage in animals (which ones?),” he wrote on February 4 to Farrar and Fauci. He doubtless meant that the virus could have been put in serial passage between animals and then escaped in some accident.

“?? Serial passage in ACE-2 transgenic mice,” Fauci replied. Fauci was suggesting that SARS-CoV2 could have acquired its surprisingly strong affinity for the ACE2 protein that studs the surface of human lung cells if it had been grown in mice genetically engineered to carry this human protein on their lung cells. Such mice stand in for human volunteers in testing whether a genetically engineered virus can infect people. The Wuhan Institute of Virology possessed a colony of ACE2-carrying mice.

“Exactly!” Farrar replied to Fauci’s suggestion.

Collins clearly found it astonishing that such a dangerous experiment would be done in low safety conditions.

“Surely that wouldn’t be done in a BSL-2 lab?” he asked Farrar.

“Wild West . . . ,” was Farrar’s response.

The Wuhan researchers did indeed manipulate viruses and cell cultures in BSL-2 labs, the second-lowest of the four designated safety levels, and hardly more stringent than the conditions in a dentist’s office. For infecting transgenic mice with engineered viruses, they used BSL-3 conditions, which are substantially safer, though insufficient to have prevented four escapes of the SARS1 virus from a lab in Beijing in 2004.

Fauci by this time knew that his agency was funding the Wuhan Institute of Virology to manipulate coronaviruses. It is unclear how much of this he had told Collins. A possible interpretation of the emails is that they reflect a process stage-managed by Farrar and Fauci to accomplish two ends. One was to supply evidence to explain why the American virologists were recanting their initial conclusion about SARS-CoV2 being lab-made. The other was to persuade the two senior officials, Collins and Vallance, that there was nothing to see here, just a little misunderstanding that could safely be put to rest.

What single piece of evidence could best support the idea that SARS-CoV2 had moved naturally from bats to an intermediary animal host and then to people just as did its predecessor, the SARS1 virus that caused the 2002 epidemic? Obviously, a coronavirus that was currently infecting some animal population and possessed a genome sequence extremely similar to that of SARS-CoV2. A virus that differed by less than 1 percent would be a plausible ancestor, strongly implying that SARS-CoV2 had emerged from nature. And Farrar had just such a surprise to pull out of his hat at the right moment.

“Reports coming out overnight that Chinese group have pangolin viruses that are 99% similar,” Farrar emailed the others on February 7. “This would be a crucially important finding and if true could be the ‘missing link’ and explain a natural evolutionary link,” he explained. It would indeed. If true.

It wasn’t. But Collins and Vallance lapped up the exciting news. “Has the actual sequence of the pangolin coronavirus isolate been released?” Collins replied, the same day. “That will be VERY interesting. Does it have the furin cleavage site?”

The next day Vallance, the U.K.’s chief scientist, replied with advice “to make sure that the sequence data from the pangolins is included” in the paper under preparation.

Both Farrar and Holmes have close connections with Chinese health officials. Could they have asked their Chinese colleagues for any data helpful to the cause? Whether or not there was any such coordination, the Chinese needed no encouragement to help discredit the idea of lab leak.

“The strain isolated from pangolin is 99% similar to the new coronavirus strain,” a perfectly timed press release from the South China Agriculture University reported on February 7, just as Farrar had predicted.

The pangolin data were highly uncertain and preliminary. But why subject it to zealous scrutiny when it supported the conclusion everyone now wanted? “Personally, with the pangolin virus possessing 6/6 key items in the receptor binding domain, I am in favour of the natural evolution theory,” Holmes wrote in an email of February 8.

On that basis, further discussion seemed unnecessary. The draft of the Proximal Origins paper was wrapped up and posted online ten days later. The colossal embarrassment of discovering that Chinese researchers funded by the NIAID had unleashed the deadly Covid epidemic on the world was averted. What was not to like?

Nothing except that the pangolin data that had driven the whole process were a gigantic red herring. Far from being 99 percent similar to SARS-CoV2, the coronavirus found in pangolins was in fact even less similar than RaTG13, a coronavirus sequence the Chinese had already published. The virus contained no furin cleavage site, so had no relevance to the principal anomaly in the SARS-CoV2 sequence. The pangolin data were not new, as Farrar had implied to Collins and Vallance, but had been made available on the website on January 23. And for all the weight placed on them by the email participants, the pangolin data had many inconsistencies. One of the preprints based on them was never published in a journal. The other was posted by the Chinese on February 20, only after the text of Proximal Origins was in near-final form. This was the unverified data used to dismiss the quite substantial evidence already available in favor of the lab-leak hypothesis. And all the email participants just went along with it.

As for the Proximal Origins paper, it was submitted to Nature, which rejected it, apparently for the hilarious reason that the European virologists were not given sufficient credit for their part in disabusing the American virologists of their belief in the lab-leak thesis. The manuscript was then passed down to Nature Medicine, a subsidiary journal.

One other point is worth noting before addressing the central mystery of the new emails. Inside the anomaly of the furin cleavage site is another puzzle, also highly indicative of an engineered virus. The genetic code is universal but also loose enough to allow for spelling preferences that differ from one organism to another. So coronaviruses prefer one set of spellings and humans another. Six of the 12 nucleotides in the furin cleavage site, the sequence CGG-CGG, represent the human-preferred spelling. Indeed, this sequence, when in correct frame, is unknown in coronaviruses, raising the clear possibility that it came from a lab kit, not from nature.

Strangely, this salient point is not even discussed in the emails or the Proximal Origins paper. Could Andersen and Holmes, both expert virologists, really have failed to notice it? Perhaps, because by February 4 Andersen was already calling the lab leak a “crackpot” theory, and as early as February 2 he may have been focused on disproving the lab-leak hypothesis, as he states in the emails, not on looking for pertinent evidence.

This brings us to the overriding puzzle of the email record. What exactly happened on February 1 to make Andersen and Holmes change their minds 180 degrees? It was not scientific evidence, because no new evidence is cited in the emails on that date. So it’s reasonable to consider other possible sources of persuasion.

Go back to the January 31st email in which Andersen tells Fauci that if you “look really closely at all the sequences” of SARS2 and related bat viruses, you can see “that some of the features (potentially) look engineered.” Hence, he wrote, he and his three colleagues, Holmes, Farzan, and Garry, “all find the genome inconsistent with expectations from evolutionary theory,” meaning it wasn’t made in nature.

This would be a stunning discovery, if true, and one that its authors doubtless intended to publish as soon as possible, even though it would trigger a political maelstrom. Andersen surely expected praise for this sensational insight and that Fauci would give helpful advice about presenting the discovery to the public.

But would the NIH leadership indeed have been overjoyed to see the mother of all public inquiries into the possibility of a link between its support of research at Wuhan and the outbreak of the pandemic? Quick action would be needed to avert any such disaster. Instead of appreciation, Andersen finds himself the very next day on a teleconference with Fauci, Collins, and senior British medical officials, being confronted by a bunch of European gain-of-function virologists baying for his blood. To Andersen’s dismay, it seems that Fauci is not so pleased. Collins, too, is seriously unhappy, fretting in an email the next day that “the voices of conspiracy will quickly dominate, doing great potential harm to science and international harmony.”

Evidently, Andersen has gotten the wrong answer. Right scientifically, maybe, but oh, so erroneous politically.

Being wrong was no trivial matter for Andersen’s group. Fauci, Collins, and Farrar between them control a huge chunk of the money available for virology research. What serious prospects could there be for Andersen’s future career if he should persist, despite heavy hints, in doing “great potential harm to science and international harmony?” As to the likelihood of NIH delight in his group’s landmark discovery that SARS-CoV2 bore the fingerprints of human design, the scales that day fell from his eyes.

“I see that road to Damascus conversion occurring during the February 1 teleconference,” Richard Ebright, a molecular biologist at Rutgers University who has long criticized gain-of-function virology research, said in an interview. “They felt they had found something really important and their funding agency would be proud of it too, but on the teleconference they learned the opposite. They felt Fauci would really want to know, but it’s the last thing he wanted to know.”

Telling the truth had landed them in deep ordure. “They were so concerned they had damaged their fundability that the only way they could retrieve it was to suborn science on behalf of their paymaster,” Ebright avers.

The Proximal Origins paper was the recantation of the dangerous conclusion they had outlined in their January 31st email. Only when this palinode is safely in press does Andersen receive the praise he had at first expected, though for the opposite reason. “Nice job on the paper,” Fauci writes him on March 8. And what a pleasant surprise: on May 21, Andersen’s lab, with Garry as a subcontractor, was awarded a grant of almost $1.9 million from Fauci’s agency.

“Our analyses clearly show,” wrote the authors of Proximal Origins, “that SARS-CoV-2 is not a laboratory construct or a purposefully manipulated virus.” For the following two years, this aggressively stated conclusion in a high-visibility journal became accepted wisdom throughout the world. The new emails give no reason for confidence that its authors believed a word of it.

Photo: Furtseff/iStock


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next