It’s clear that SARS-CoV-2, the coronavirus behind the COVID-19 pandemic, is most closely related to a group of viruses that usually infect bats. But exactly how and where it evolved to become such an efficient respiratory pathogen remains to be seen. Now, in a study published July 9 in Nature Structural & Molecular Biology, researchers have determined that the spike proteins of SARS-CoV-2 and of the closely related bat coronavirus RaTG13—while similarly structured overall—differ in their stability and affinity for binding ACE2, the receptor that SARS-CoV-2 uses to infect human cells. 

The substantial difference in the spike protein of the closest viral relative “tells you that this was not a direct jump from this virus into humans,” says Amesh Adalja, a physician who studies emerging infectious diseases at the Johns Hopkins Bloomberg School of Public Health and was not involved in the work. It’s likely that SARS-CoV-2 “had been evolving in some other species—possibly an intermediate species—before it acquired the ability to be this human pathogen of such a degree that it is today.” 

A group of researchers in structural biologist Steve Gamblin’s lab at the Francis Crick Institute in the United Kingdom specializes in understanding how changes in the shapes of proteins on the surface of the influenza virus allow it to jump between different species, Donald Benton, a postdoc in the Gamblin lab, tells The Scientist. Earlier this year, when it became clear that SARS-CoV-2 was gaining steam, they decided to devote their expertise to asking the same type of questions about the coronavirus, focusing on its iconic spike proteins that protrude from the viral surface. 

Previous work showed that the SARS-CoV-2 spike protein must be cut between two amino acids at the junction between the portion of the protein that binds a receptor and the domain of the protein responsible for fusing with the host cell membrane. Rather than cutting the protein in two, this cleavage event—performed for SARS-CoV-2 by the human protease furin—is thought to increase flexibility in the protein so that it can enter mammalian cells. To investigate how this cleavage affects the structure of the protein, Benton and colleagues generated a version of the SARS-CoV-2 spike protein with the furin cleavage site intact and then exposed that protein to furin to generate a cleaved version. 

In the uncleaved form, the protein was stable, with three components known as receptor-binding domains (RBDs), which are thought to bind ACE2, tightly tucked into the top of the protein. After furin cleavage, one of the RBDs rotated to open a surface at the top of the protein for ACE2 interaction. The findings indicate that furin cleavage appears to make the spike protein more likely to adopt an open shape that allows it bind to the receptor and enter human cells. 

According to a study published in February, SARS-CoV-2 and RaTG13 share about 96 percent of their genomes and about 93 percent sequence similarity in their spike protein genes, making RaTG13 the closest SARS-CoV-2 relative found yet. In work published in April, researchers showed that the amino acid sequences of the two proteins were least similar—around 90 percent—in the RBDs and that the furin cleavage site in SARS-CoV-2’s spike protein is absent in RaTG13, findings Benton and colleagues confirmed. 

The authors of the new paper also observed that the SARS-CoV-2 spike protein binds ACE2 about 1,000 times more tightly than the RaTG13 spike protein.

“It looks as if this particular bat virus wouldn’t directly be able to infect humans because of its weak ability to bind to the human receptor,” Benton tells The Scientist. 

“We probably still haven’t found the correct bat virus that actually did make this leap” to people, he says, though there are some coronaviruses in pangolins—scale-covered mammals found in Asia and Africa—that have similar RBDs in their spike proteins. 

In a study published July 1, researchers proposed that recombination events between multiple coronaviruses from different species—potentially including RaTG13—could have led to the emergence of SARS-CoV-2, an idea that’s supported by findings of the current study, according to Benton and his colleagues. “That’s not just plausible, I think it’s also parsimonious,” says Adam Frost, a structural biologist at the University of California, San Francisco, who did not participate in either study. “Where that recombination event took place remains unknown . . . and it may be very hard to be absolutely sure.”

Beyond the evolutionary insight, these new structures may also help researchers generate tools—such as antibodies and synthetic ACE2 domains—that could attach to the RBD and prevent it from engaging the endogenous ACE2, Frost explains. “Big picture, these new structural states will help us both develop and understand those kinds of therapeutic reagents.”