Jason McLellan was wandering around a ski shop of Utah’s Park City Mountain Resort, waiting for his new snowboarding boots to be heat-molded to his size-nine feet, when his smartphone rang. It was Barney Graham, deputy director of the National Institute of Allergy and Infectious Diseases Vaccine Research Center.
Two days earlier, the World Health Organization had announced that several unidentified pneumonia-like cases had been reported in Wuhan, China. People were fatigued and feverish, with dry coughs and headaches. These symptoms weren’t unusual for early January, but some people were short of breath, and a few felt like they’d been hit by a train.
Graham told McLellan, a structural virologist at the University of Texas at Austin, that the ailment appeared to be a beta-coronavirus, meaning it fell into the genus of viruses that causes severe acute respiratory syndrome (SARS). He asked McLellan: “Are you ready to get back in the saddle?”
This duo was part of a small band of government and university scientists who had spent more than a decade cracking a complex viral puzzle—and their skills were needed once more. Their years of sleuthing and innovating ultimately contributed a microscopic but critical piece to the most promising candidates for COVID-19 vaccines. Two already authorized in the U.S. use their discovery, as do at least two other top contenders.
Their solution? Tweaking a shape-shifting protein to make it sit still.
Stabilizing the trickster
By the time McLellan landed in 2008 at the Vaccine Research Center in Bethesda, Maryland as an early-career researcher, Graham had been working on a little known but highly contagious disease caused by respiratory syncytial virus for more than 20 years. Both the cold-causing RSV and the SARS-CoV-2 coronavirus, which causes COVID-19, feature genomes made of RNA. Although the two sit on distant branches of the evolutionary tree, they share a common physical trait that would yield the first key to McLellan and Graham’s journey toward beating COVID-19.
Attempts to design an RSV vaccine had been riddled with hiccups since 1966 when a clinical trial inadvertently enhanced the illness in volunteers—and even caused the death of two infants. Graham wanted to understand why this drug candidate had failed so terribly.
Similar frustrations hovered around another germ under study at the Vaccine Research Center: HIV. McLellan had arrived at the center to train with Peter Kwong, a structural biologist tinkering with the structures of viral proteins in the hopes of engineering a vaccine that would stop AIDS. HIV rapidly mutates, so the researchers tried several structural biology tricks to develop vaccine candidates but ultimately failed to create one that elicited an immune response.
“You didn’t know whether it was because the virus was too good or the ideas were bad,” McLellan says.
In what the pair now refers to as a happy accident, Graham and McLellan were working near one another on the center’s second floor. Kwong’s fourth-floor lab was too crowded for McLellan, so he set up a workspace within earshot of Graham, and they became friends. “It didn’t take long for him to come to me and say, I’d like to work on something other than HIV,” Graham recalls.
Past unsuccessful attempts to neutralize RSV with a vaccine had focused on the virus’s class 1 fusion protein, or F protein. In the wild, this protein is a shapeshifter, “like a Transformer toy,” Graham says. It can look one way before the RSV virus infects and enters a cell, and another way after the virus multiplies and escapes. These Jekyll-and-Hyde identities are known as the “prefusion” and “postfusion” states, and all vaccine attempts up until this point had focused on the latter.
To make matters trickier, the prefusion form is extremely unstable: It can irreversibly and spontaneously snap to its other state in an instant. Graham and McLellan hypothesized that they might create a more successful RSV vaccine if they could lock in the prefusion state. But no one knew what the prefusion protein looked like; they just knew it was a trickster.
So, McLellan used x-ray crystallography—a technique that uses x-ray beams to determine the structure of proteins—to capture an image of the prefusion protein for the first time. Some researchers would later say the prefusion F protein looked like a lollipop. McLellan thought it looked like a Nerf football. “You’re one of the first people in the world to see what this protein looks like,” he says. “It’s pretty cool.”
By examining the protein at this atomic level, McLellan found a way to bioengineer it to take away its shape-shifting power. In other words, he stabilized it.
When Graham tested this new molecule in animals, it acted as an antigen and stimulated the immune system to fight disease. It had 50 times more neutralizing power against RSV than anything he had tested before. On the flip side, they also showed a postfusion version of the protein takes on an identity that can bypass the immune system’s defenses.
Their accomplishment won runner-up recognition in Science’s 2013 Breakthrough of the Year, and their work carved the way for new RSV vaccines that are showing great promise, Graham says.
“The work of Jason and Barney and others revolutionized the field,” says Ruth Karron, a professor of international health at Johns Hopkins Bloomberg School of Public Health and the director of the Center for Immunization Research and the Johns Hopkins Vaccine Initiative.
The lucky last step
Five years ago, a postdoctoral fellow in Graham’s laboratory returned from a trip to Saudi Arabia with a respiratory infection. Everyone assumed the fellow had Middle East respiratory syndrome (MERS), caused by a dangerous coronavirus that had arisen in the country two years earlier.
That emergence happened around the same time that McLellan launched his own lab at Dartmouth College in New Hampshire. McLellan and Graham had been trying the prefusion trick on MERS, given that coronaviruses feature spike proteins that are also shapeshifters and are used to break into our cells. When Graham’s lab tested the postdoc’s nasal secretions, they found a related germ—and an opportunity that would pave their final steps toward a COVID-19 vaccine.
The postdoc had an older coronavirus: HKU1, a mild cold-causing bug that was discovered in 2005. The Graham-McLellan partnership decided to pivot their focus to HKU1 because MERS required extra safety precautions, and their research on the latter had hit a wall.
To capture a 3D picture of HKU1, McLellan would need a different method for taking atomic-level pictures. X-ray crystallography saturates proteins in a salt bath solution until they form crystals akin to rock candy. But due to their physical nature, coronaviruses don’t crystalize well. Cryogenic electron microscopy, or cryo-EM, is a technique that allows scientists to view proteins frozen in a thin layer of ice, bypassing the need for crystallization.
Proteins are so small that you can’t use a regular light microscope to take a picture. Scientists used a cryo-electron microscope to determine the SARS-CoV-2 spike’s structure. Cryogenic electron microscopy is a technique that allows the researchers to view proteins frozen in a thin layer of ice using beams of electrons. The proteins are lying in a variety of orientations, leaving an assortment of shadows. The scientists then combine all these 2D images of shadows to create a 3D shape.
In 2015, structural biologist Andrew Ward was one of the leading cryo-EM experts in the U.S., so McLellan emailed his lab at Scripps Research in San Diego to ask if he had any interest in studying coronaviruses. Coincidentally, Ward had a postdoctoral fellow with a hankering to examine coronaviruses. They ultimately took thousands of images of HKU1 proteins.
McLellan used this 3D readout of HKU1 to make educated guesses at how to stabilize the spike proteins from its viral cousins, MERS and SARS. McLellan and Nianshuang Wang, his postdoctoral fellow, discovered that by adding two prolines—rigid amino acids—to MERS’s spike protein, they could prevent it from changing shape.
They called the tweak a 2P mutation and filed a patent for it in 2017. Around the same time, Graham’s lab partnered with biotech company Moderna to design an experimental mRNA vaccine for MERS. The two had worked together a year prior on a similar but separate project to combat the Zika virus—as part of a new movement for more comprehensive preparations against global outbreaks. The concept hinged on the detailed study of a prototypical member of a viral family—such as HKU1 or MERS—to build defenses against all future troublemakers from the same family like SARS-CoV-2.
Ultimately, experiments in animal models showed the MERS vaccine was successful, says Kizzmekia Corbett, a postdoctoral research fellow in Graham’s laboratory, and created a “portfolio of data” that the scientists knew they could apply to the new coronavirus.
The road to salvation
On January 6, 2020, just minutes after he took that phone call at the ski shop, McLellan messaged Wang and Daniel Wrapp, a graduate student, on WhatsApp.
“Barney is going to try and get the coronavirus sequence out of Wuhan, China,” McLellan wrote to them. “He wants to rush a structure and vaccine. You game?”
The two labs worked in concert with one another, determining the virus’s structure in about two weeks and using the 2P mutation to stabilize its proteins. Graham’s lab partnered with Moderna, and Corbett designed and executed clinical assessments to immunize mice with an mRNA vaccine made with the modified proteins starting in February. “When we got the first results from the mice, and they had a great antibody response, it was so gratifying,” Corbett says. By March 4, the U.S. Food and Drug Administration had greenlit the Moderna vaccine for human trials.
At about the same time, Pfizer and BioNTech spoke with Graham about using the 2P mutation in their vaccine. Because their work was patented and widely published, other drugmakers—including Novavax and Johnson & Johnson—also based their candidates on the design. Pfizer-BioNTech’s vaccine would become the first authorized in the U.S. after it showed an impressive 95-percent efficacy rate. Moderna’s vaccine was 94-percent effective.
Further tests would be needed to judge how much the 2P mutation contributes to the overall efficacies of the frontrunner vaccines. Phil Dormitzer, Pfizer’s chief scientific officer and vice president of viral vaccines, says it’s “absolutely clear” that stabilizing prefusion proteins led to remarkable advances with potential RSV vaccines. “I’m very glad we picked those mutations to move forward,” he says, referring to the Pfizer-BioNTech COVID-19 vaccine.
Graham doesn’t quite know how to answer when asked how it feels to have decades’ worth of work contribute to rapidly developed vaccines that could save hundreds of thousands of lives amid a harrowing global pandemic. “That’s not the way we usually think about it,” he says. “I don’t think you really think that much about your feelings until you get to certain milestones.”
But the question—posed using the phrase “such a time as this”—makes Graham hearken back to the biblical tale of Esther, a queen who was made a royal for “such a time as this.”
“I have kind of felt like my whole career has been lining up for ‘such a time as this,’” Graham says.