The Story of the Decade: New documents strengthen—perhaps conclusively—the lab-leak hypothesis of Covid-19’s origins.
The day is growing ever closer when Washington may have to add to its agenda with Beijing a nettlesome item it has long sought to avoid: the increasingly likely fact that China let the SARS2 virus escape from the Wuhan lab where it was concocted, setting off the Covid-19 pandemic that killed some 7 million people globally and wrought untold economic havoc.
New documents may explain why no one has been able to find the SARS2 virus (aka SARS-CoV-2) infesting a colony of bats, from which it might have jumped to people. The reason would be that the virus has never existed in the natural world. Documents obtained by U.S. Right to Know, a health advocacy group, provide a recipe for assembling SARS-type viruses from six synthetic pieces of DNA designed to be a consensus sequence—the genetically most infectious form—of viruses related to SARS1, the bat virus that caused the minor epidemic of 2002. The probative weight of the recipe is that prior independent evidence already pointed to SARS2 having just such a six-section structure.
The documents unearthed by U.S. Right to Know, and analyzed by its reporter Emily Kopp, include drafts and planning materials for the already-known DEFUSE proposal, an application to DARPA, a Pentagon research agency, for a $14 million grant to enhance SARS-like bat viruses.
The new recipe is in striking accord with a theoretical paper published in 2022 that predicted the SARS2 virus had been generated in exactly this way. Three researchers—Valentin Bruttel, Alex Washburne, and Antonius VanDongen—noted that the virus could be cut into six sections if treated with a pair of agents known as restriction enzymes and so had probably been synthesized and assembled in this way.
Restriction enzymes, made naturally by bacteria as a defense against viruses, are an invaluable tool for biologists because they cut DNA at specific points known as recognition sites. These sites occur randomly across the genome, so a natural virus treated with a restriction enzyme will be cut into pieces of different sizes. However, researchers who want to synthesize a virus from scratch in order to manipulate its parts more effectively will often rearrange the recognition sites so that they are evenly spaced. This allows short chunks of DNA, all of roughly equal length, to be synthesized chemically and then strung together in a complete viral genome. Bottom line: if your virus has evenly spaced recognition sites, it’s a pretty good bet that it was made in a laboratory.
Bruttel and his colleagues guessed that a commonly used pair of restriction enzymes, known as BsaI and BsmBI, might have been used to assemble the SARS2 virus’s genome. When they examined the structure of SARS2, they found that the recognition sites used by these enzymes were indeed evenly spaced across the genome, marking it into six sections. “Our findings strongly suggest a synthetic origin of SARS-CoV2,” they wrote.
Their paper did not receive the attention it deserved, in part because of the difficulty of ruling out a natural explanation for the even spacing. The small group of virologists who adamantly oppose the lab-leak hypothesis attacked the paper as “confected nonsense” (Edward Holmes) and “kindergarten molecular biology” (Kristian Andersen).
The recipe in the new DEFUSE drafts, however, closely resembles the one posited in the Bruttel article in saying that new viruses would be constructed from six sections of DNA synthesized in a lab. The documents even include a form for ordering the BsmBI restriction enzyme.
“The DEFUSE draft documents show that, exactly as we had postulated, they planned to use 6 segments to assemble synthetic viruses, to use unique endonuclease sites that do not disturb the coding sequence, and TO BUY BsmBI !!!” Bruttel wrote in a post on X.
The fact that the rearranged recognition sites change the virus’s nucleic-acid sequence but not the proteins it specifies is unusual and would occur by chance less than one in a million times, Bruttel said in a lecture.
His coauthor Washburne said in an email, “For us, it’s not a surprise to find DEFUSE collaborators using these enzymes & discussing a 6-segment assembly because the odds of this pattern occurring in nature have been very low this whole time.”
Discovery of the new recipe certainly strengthens the possibility that the regular spacing of BsaI and BsmBI recognition sites in SARS2 is the signature of synthetic origin. Indeed, Richard H. Ebright, a molecular biologist at Rutgers University who had called the 2022 paper “noteworthy . . . but not decisive,” now says that the evidence in the new documents “elevates the evidence provided by the genome sequence from the level of noteworthy to the level of a smoking gun.”
Some experts say that the sequence issues highlighted in the 2022 paper and the new DEFUSE documents need further study before they can be claimed as definitive. Others think that the claim can be made already.
“Game over,” wrote Matt Ridley, co-author of Viral: The Search for the Origin of COVID-19, noting that every suspicious feature of SARS2 is explained by the methods called for in the DEFUSE proposal documents. But his coauthor, Alina Chan, argues that more needs to be understood about what happened after 2018.
The DEFUSE project, first leaked in 2021, was submitted in 2018 but turned down by DARPA. That doesn’t mean that the experiments it describes were not performed. It’s common practice to strengthen a grant application by doing much of the proposed work beforehand. Or the researchers may have found funds elsewhere.
The DEFUSE proposal was authored by Peter Daszak, head of the EcoHealth Alliance in New York, with partners including Shi Zhengli of the Wuhan Institute of Virology and Ralph Baric of the University of North Carolina. The grant proposed to “introduce appropriate human-specific cleavage sites” into SARS-related viruses, a procedure that could have led to the creation of SARS2, with its distinctive furin cleavage site, depending on the starting virus used for the manipulation.
The new drafts show the authors planned to synthesize eight to 16 strains of SARS-type bat viruses, selected for their likely ability to infect human cells. The goal was to use them to make a vaccine to immunize bats in regions that military troops might have to enter. The researchers were well aware of the risk that their work would set off a pandemic. “Also, we MUST make it clear in proposal that our approach won’t drive evolution the wrong way (e.g. drive evolution of more virulent strain that then becomes pandemic,” says a planning memo.
Some observers believe that when DARPA declined to fund the project, the Chinese members of the group may have decided to find their own financing and go ahead unilaterally. This is plausible, as Baric and Shi were collaborators but also rivals. With Baric blocked for lack of DARPA funds, Shi may have seen the chance to race ahead if she could acquire funds from Chinese sources.