In mid-2020, a team of scientists catching bats in Laotian caves discovered coronaviruses that were strikingly similar to the one that had begun wreaking havoc around the world.
In the months since, some of those researchers have been studying one of these mysterious bat viruses in a high-security laboratory in Paris, hoping to discover clues about how its cousin, SARS-CoV-2, went on to become a global threat that has killed an estimated 15 million people.
Their work has been scientifically fruitful. Last year, the scientists discovered that the bat virus was capable of latching on to human cells, at least in Petri dishes. Last month, the team reported more reassuring news: that the virus is not particularly harmful to lab animals. The finding suggests that SARS-CoV-2 evolved its abilities to spread quickly and cause deadly disease only after the two lineages branched apart on the viral evolutionary tree.
If the Laotian virus were to ever jump from a bat to a person, the new research suggests, it might cause a mild stomach bug rather than a life-threatening pneumonia. Nevertheless, lab experiments like these rekindle a longstanding debate among scientists about the wisdom of tinkering with viruses that are so closely related to a known pathogen.
Proponents argue that this kind of data is crucial for understanding — and preventing — pandemics. For example, the new studies have tested whether bat viruses could evolve a “furin cleavage site,” a feature of SARS-CoV-2 that allows it to efficiently infect human cells.
“Our motivation was to try to give some insight regarding the origin of Covid,” said Marc Eloit, a virologist at the Pasteur Institute in Paris who is leading the effort.
But critics say that scientists should not run experiments that might make viruses better able to spread among people, given the small but real chance that these altered pathogens might infect lab workers and escape into the outside world.
“For me, the benefits of this work are outweighed by the risks,” said Dr. David Relman, a microbiologist at Stanford University.
The furin site
The bat virus at the center of Dr. Eloit’s experiments came to light on an expedition to limestone caves in northern Laos in the summer of 2020. A team of Laotian and French researchers caught bats flying out of the caves and took samples of their saliva, blood, urine and feces.
The scientists found genetic material from five coronaviruses closely related to SARS-CoV-2. In the feces from a Marshall’s horseshoe bat, they found whole viruses of a strain they named BANAL-236 (a code indicating the virus came from a bat anal swab).
Once back in their lab, the scientists found that BANAL-236 can infect human cells by binding tightly to the same protein that SARS-CoV-2 uses to gain entry. In February, the researchers published these findings in the journal Nature.
Last month, they released a second wave of results, now under review by a scientific journal, in which they investigated the virus’s behavior in laboratory mice and monkeys.
In one experiment, the scientists injected the virus into genetically engineered mice commonly used to study Covid. SARS-CoV-2 replicates quickly in their lungs, just as it does in people, causing them to lose weight and die.
BANAL-236, by contrast, struggled to take hold in the animals’ lungs, producing only about 1 percent of the viruses created in a SARS-CoV-2 infection.
The researchers found that the virus was even milder when they sprayed it into the noses of two monkeys. BANAL-236 replicated mainly in their guts, rather than their lungs.
Dr. Eloit suspects BANAL-236 is milder because it lacks a key feature important to the success of SARS-CoV-2.
After a new SARS-CoV-2 virus is created in a cell, its spike protein changes shape, with an effect like spring-loading a crossbow. When the virus then binds to a new cell, the primed spike protein shoots out molecular bolts that draw it into its new host.
This shape-shifting region of the spike — known as the furin cleavage site — is crucial to the success of SARS-CoV-2. When scientists have engineered viruses lacking this site, the mutants struggle to replicate in the lungs of lab animals or spread to new hosts.
Gaining a furin cleavage site may have been a crucial step in the evolution of SARS-CoV-2. To explore that possibility, Dr. Eloit and his colleagues ran lab experiments to give BANAL-236 a chance to evolve new traits, such as a furin cleavage site.
The team based its studies on experiments in which other scientists had injected bird flu viruses into chicken eggs and waited for them to replicate. They then transferred the new viruses into new eggs, and again allowed them to replicate. With every transfer, the virus had a chance to evolve. After 11 transfers, the scientists found that the flu viruses had evolved cleavage sites, making them deadlier to chickens.
In a similar fashion, the Pasteur researchers removed lung tissue from mice infected with BANAL-236 and used the tissue to infect healthy animals. They then repeated the cycle, transferring viruses from mouse to mouse.
In another experiment, they infected a dish of human intestinal cells with BANAL-236, then used the new viruses produced by the cells to infect new dishes.
But for both experiments, Dr. Eloit and his colleagues decided against going as far as 11 transfers, stopping at six.
“From a purely scientific point of view, we wanted to do more than six passages,” Dr. Eloit said. “But we did not want to open the risk to adapting a bat virus to humans.”
BANAL-236 did not gain a furin cleavage site in either experiment. The virus did acquire other mutations, but they did not make it any better at infecting mouse lungs.
Scientists have been running such evolutionary experiments — known as “serial passage” — for more than a century. In fact, vaccines for a number of viruses such as yellow fever were created by growing them in the lab: The viruses evolved in Petri dishes into milder forms that were safe to inject into people.
In 2011, however, a controversy broke out about the safety of serial passage experiments that might produce new human pathogens. At the time, researchers were studying how influenza viruses that cause intestinal infections in birds can evolve into airborne forms that can infect people.
Two teams of researchers sprayed bird flu viruses into the noses of ferrets, waited for the viruses to replicate and then transferred the new viruses to new ferrets. Soon the viruses evolved to become better at replicating in the ferrets.
Some critics said the research was so reckless that it shouldn’t be published, for fear that other researchers would copy the work and accidentally release a new pandemic strain of flu. The United States government halted experiments like these in order to develop a new policy for judging their safety.
Some of the studies started up again in recent years. But Dr. Relman of Stanford and others have complained that the current regulations are not transparent enough.
Dr. Eloit said a Pasteur Institute committee that reviews potentially risky biological research authorized his team’s proposal for studying the new bat viruses. The scientists then carried out their experiments at the same level of security as their other work with the coronaviruses, known as Biosafety Level 3, or BSL-3.
Dr. Tom Inglesby, the director of the Johns Hopkins Center for Health Security at the Bloomberg School of Public Health, said it was good that the scientists put thought into these potential risks. But he also said he wanted to see a clear rationale for deciding that six passages were safe.
“It’s not possible to know ahead of time if these experiments would lead to more transmissible or more virulent viruses,” he said. “There’s no hard and fast rule that six is safe and more is not.”
But Thomas Peacock, a virologist at Imperial College London, said he thought that Dr. Eloit and his colleagues had been sufficiently prudent. In earlier studies, he pointed out, the researchers had found that antibodies produced by people during Covid infections were very potent against BANAL-236. That most likely meant that if the virus were to leak out of a lab, it wouldn’t be able to spread very far.
“This virus would probably hit a brick wall in the general population,” Dr. Peacock said. “I don’t really have much issue with the experiments.”
Origins of Covid
Other researchers agreed with Dr. Eloit that the research could shed light on how and when SARS-CoV-2 spilled over into people.
For Dr. Eloit, his team’s inability to produce a furin cleavage site on BANAL-236 in mice or human intestinal cells suggests that the SARS-CoV-2 lineage gained the furin site in bats before spilling over into people. He said it would not have been easy for the virus to gain a furin site after jumping into another species of animal — sometimes called “intermediate hosts” — such as those sold at a market in Wuhan, China. “I do not see any strong arguments in favor of an intermediate host,” Dr. Eloit said.
But scientists who favor the market scenario see the new results in a different light. If the researchers couldn’t spur BANAL-236 to evolve the furin site during serial passage experiments, they reasoned, then it’s unlikely that scientists in a Wuhan lab could have done so with SARS-CoV-2, as some proponents of the “lab leak” theory have suggested.
“This is another nail in the coffin of the lab leak theory that by now should be firmly sealed in the crypt,” said Edward Holmes, a virologist at the University of Sydney.
Dr. Peacock was reluctant to draw strong conclusions from such small-scale experiments. “I think it’s quite a difficult thing to ask to get a furin site after a few passages,” he said.
Dr. Eloit and his colleagues are now exploring the possibility that ancestors of SARS-CoV-2 gained a furin cleavage site while still in wild bats. The virus might have then spread to an intermediate host or directly to people exposed to bats — like those who collect bat guano, hunt bats or eat them.
To test that idea, the scientists are working to obtain more samples from bats in Laos and neighboring countries. Dr. Eloit can’t say whether their hypothesis is more likely than the others, but it is at least one they can investigate.
“Our work as scientists,” he said, “is to explore the working hypotheses that we can explore.”