Researchers have created the most complete atomic-level simulation to date of the SARS-CoV-2 spike protein, which the virus uses to enter host cells. The model reveals that the glycans don’t just camouflage the virus from the host immune system, as sugars on proteins are known to do. The glycans on SARS-CoV-2 also help the spike protein take on the conformation that equips it to infect cells (ACS Cent. Sci. 2020, DOI: 10.1021/acscentsci.0c01056).
“We have an atomic-level model of a fully glycosylated, full-length spike protein, so there’s no more missing bits,” says Rommie Amaro, a computational biophysical chemist at the University of California San Diego. And by looking at the model, she explains, “we realized that a couple glycans in particular were not only acting as part of the shield, camouflaging the virus, but they also seemed to be acting as part of the weaponry.”
More than 8 months into the global COVID-19 pandemic caused by SARS-CoV-2, researchers around the world have generated multiple detailed images of the spike protein using a technology called cryo-electron microscopy (cryoEM). But one of the limitations of cryoEM is that it can’t resolve bits of the protein that are very flexible, such as the glycans that cover the virus’s outer surface, including those covering the spike protein. Glycans “are like little leaves that hang off branches, which are the protein residues,” Amaro explains. CryoEM can see where these branches attach to the spike protein itself, and most of the branches, but the leaves themselves are invisible to cryoEM.
To better visualize the glycans, Amaro and her colleagues used a computational technique to combine multiple existing cryoEM models of the spike protein with biochemical analyses that indicate the types of glycans that are present and where along the protein they might lie. From all this information, they created a dynamic movie of how all the atoms in the protein “wiggle and jiggle over time,” she explains.
They noticed two striking things about the glycans: they were unusually numerous, and the behavior of two particular glycans seemed to play a special role. In order to infect a cell, the spike protein must first undergo a set of conformational changes. Specifically, its receptor binding domain—a spear-shaped region near the end of the spike—must raise itself upwards toward the host cell receptor to make the connection. The two glycans appeared to tuck themselves under this domain and help lift it toward the receptor in order to infect cells.
To determine whether the glycans did in fact perform this job, they collaborated with Jason McLellan at the University of Texas at Austin, whose team was the first to determine the structure of the SARS-CoV-2 spike protein using cryoEM in February. McLellan and his colleagues deleted the two glycans by mutating the amino acids they attach to on the spike protein. Lab testing showed that compared to the unmutated spike protein, the mutated proteins were less efficient—by half for one mutant—at interacting with host cell receptors.
Glycans are well known to shield or camouflage proteins from the immune system, as well as to help proteins fold, Amaro explains. But this study is among the first to suggest that glycans can directly change a protein’s conformational dynamics. It may be possible to develop therapeutics that target the glycan processing pathway, Amaro says.
The work provides “a tour-de-force, all-atom model construction and molecular dynamics simulation,” says Tamar Schlick, a chemist and computational scientist at New York University. The data indicate that glycans play an essential role in the spike protein’s binding to the host cell receptor, she says.
Amaro and her colleagues are now using their technique to look more closely at the interaction between the spike protein and the host receptor. “Between the first contact with the human cell and the engulfment of the virus by the cell there’s a huge set of conformational changes that happen,” she says. Those, too, may offer targets for therapeutics or vaccine efforts.