Spring 2022


This image, constructed from data gathered through cryogenic electron microscopy (cryo-EM), shows the substitution of two amino acids called prolines (in purple) into the SARS-CoV-2 spike protein. The innovative modification set the stage for improved imaging of the virus and spurred the creation of coronavirus vaccines. Image courtesy of Robert Kirchdoerfer


It was a quiet evening in 2012, and Robert Kirchdoerfer BS’06 was missing yet another dramatic sunset over the Pacific Ocean. But it didn’t bother him much. He had stumbled upon a riveting article, one that would guide his research for the next decade and contribute to lifesaving COVID-19 vaccines and therapeutics.

Robert Kirchdoerfer, assistant professor of biochemistry, used the Talos Arctica cryo transmission electron microscope (background) at the UW–Madison Cryo-EM Research Center to image coronavirus protein structures. Photo by Michael P. King

The paper reviewed how coronaviruses make copies of their viral genomes. At 26,000 to 32,000 bases, or nucleotides, in length, the seven coronaviruses known to infect humans have the largest known RNA virus genomes. How each coronavirus enters and replicates in cells is unique and extraordinarily complex, involving dozens of proteins and myriad processes.

Kirchdoerfer, then a postdoctoral researcher at The Scripps Research Institute in La Jolla, California, was intrigued by our limited knowledge of these viruses. Coronaviruses bring about common colds, but they also caused outbreaks of severe respiratory illnesses in the early 21st century and recently triggered the COVID-19 pandemic. As a structural biologist by training, Kirchdoerfer felt he could further our understanding by imaging virus structures (i.e., showing what virus components look like in three dimensions) and studying how these structures interact.

“After the SARS outbreak in the early 2000s, we knew a lot about coronavirus RNA synthesis structural biology,” Kirchdoerfer recalls. “We had imaged these structures, and they were supposed to be contributing to our idea of how coronavirus RNA formation worked. But we didn’t know the structure of the RNA synthesis machine itself, and I thought that was a rather glaring hole in our overall understanding.”

Research suggested that new coronaviruses would emerge in the future, so Kirchdoerfer decided to focus on how these sometimes dangerous and always enigmatic viruses ensure their survival by entering and replicating in host cells. He and his colleagues surmounted well-known challenges in structural biology over the next 10 years. What they would learn would directly contribute to the fight against COVID-19.


RNA Viruses and Vaccines
The genetic material of RNA viruses is ribonucleic acid (RNA). Human diseases caused by RNA viruses include the common cold, influenza, SARS, MERS, hepatitis C, West Nile fever, Ebola virus disease, rabies, polio, mumps, measles, COVID-19, and more. It’s difficult to make vaccines against RNA viruses in part because they generally have higher mutation rates than DNA viruses. This is because the catalyzing substances, or enzymes, which bring about the formation of polymers in RNA viruses lack the proofreading ability of those in DNA viruses.


Fortuitous Innovations

Soon after reading that life-changing article in 2012, Kirchdoerfer received a note from a friend at Scripps, Andrew Ward. Ward suggested that they work together to image coronavirus spike proteins, which are key to the virus’s ability to attach to and infect healthy cells. Though he’d originally intended to study the RNA synthesis machinery needed after a virus fuses with a host cell, the idea of exploring the spike protein (one of four external proteins in coronaviruses) was too compelling for Kirchdoerfer to pass up.

That’s in part because no one knew what complete coronavirus spike proteins looked like. Previously, scientists had passed X-rays through crystalline forms of spike proteins to create images of these structures. With this technique, called X-ray crystallography, they could only study regions of spike proteins — they didn’t know what complete spikes looked like. Also unknown at the time: how spikes interact with other virus components to trans-port coronavirus genomes into host cells.

Developments almost half a century in the making changed the game. Transformed by advances in detector technology and software, a technique called cryogenic electron microscopy (cryo-EM) could image the natural state of complete spike proteins at near-atomic scales, no crystals required.

“A lot of the strengths for looking at viruses with cryo-EM are the general strengths of the technique,” Kirchdoerfer says. “You don’t need a crystal, and it’s great for larger complexes. With cryo-EM, so much of the data is handled computationally that we can access even the moving regions of coronaviruses.”

The first coronavirus spikes Kirchdoerfer imaged came from HKU-1, a coronavirus that causes symptoms resembling the common cold but can progress to pneumonia and other respiratory illnesses. He was also eager to image the spike proteins from two public health threats, MERS-CoV and SARS-CoV. (These coronaviruses, respectively, caused outbreaks of Middle East respiratory syndrome in 2012 and severe acute respiratory syndrome in 2003.) The catch was that the spikes from MERS and SARS are unruly — they rapidly change configurations so that antibodies can’t stick to them, and they are difficult to make in a lab setting. But after one more year of hard work, inspiration struck.

Another of Kirchdoerfer’s and Ward’s collaborators, Jason McLellan, had an idea. A molecular bioscience professor at Dartmouth College at the time (now at the University of Texas at Austin), McLellan was motivated by work on other viruses, such as HIV. This prompted him and colleagues in his lab to team up with Barney Graham at the National Institutes of Health Vaccine Research Center, and together they took an innovative step in spike protein research. They substituted amino acids called prolines into the spike protein, which kept the spike in a single configuration. They also boosted the amount of the protein that can be produced in the lab. Finally, the MERS and SARS spike proteins were ready for cryo-EM imaging.


Image courtesy of the UW-Madison Cryo-EM Research Center

Intro To Cryo-EM
In cryo-EM, proteins are placed on grids that are the size of a pencil eraser. The grids are rapidly submerged in liquid ethane, which immobilizes individual particles in random orientations. The frozen grids are then placed in the cryo-electron microscope, where several hundred to several thousand images are collected. Images are turned over to computers, which isolate each particle and digitally combine them to reconstruct a 3D picture of a biomolecule. For more, see A Cold, Hard Look at Macromolecules in Grow, Spring 2021.


A COVID-19 Vaccine

By fall 2019, Kirchdoerfer had returned to his research roots, his home state, and his alma mater. He had been an undergraduate during his first stint at the Department of Biochemistry; this time, the Oregon, Wisconsin, native came back as an assistant professor. A few months after his homecoming, his spike protein structures and nascent coronavirus research lab were thrust into the spotlight.

“When we added these two prolines in the spike protein, they improved the amount of protein we got in the lab,” says Kirchdoerfer, who was first author on the resulting paper about the SARS spike protein. “In a vaccine, they could help make more stable spike proteins to raise antibodies and combat infection. That was really the push to put the modified spike protein into vaccine candidates — the antibodies that you make with a spike [that is] stabilized with prolines could make a higher antibody response than just a regular spike from infection.”

The scientists’ ingenious modification of the spike, which they had used to image spike protein structures from MERS and SARS, had also resulted in plans for a MERS vaccine with clinical trials led by Moderna. When the COVID-19 pandemic hit, these efforts were redirected to SARS-CoV-2, the virus that causes COVID-19.

“The Moderna COVID-19 vaccine was turned around so quickly because there was a vaccine for MERS already in the pipeline,” Kirchdoerfer says. “People could pull up our first SARS structures and say, ‘This is what we think SARS-CoV-2 is going to look like,’ and that was borne out by additional studies.” (The genome sequence identity for spike proteins in SARS and SARS-CoV-2 is around 80%, high enough to be similar in structure.)

Moderna, Pfizer, Johnson & Johnson, and other pharmaceutical companies use similar strategies to make COVID-19 vaccines with stabilized spike proteins as an active ingredient. Ongoing research will reveal whether substituting more prolines can make the spike even more immunogenic.

Kirchdoerfer is now known around the UW campus as “the coronavirus guy” for his expertise and collaborative nature. He synthesizes results from several state-of-the-art technologies and multiple disciplines to create a more precise map of what SARS-CoV-2 proteins look like and how they interact. Together, he and his colleagues at UW are demonstrating just how much we have yet to learn about SARS-CoV-2. Every research study, every experiment informs another, and lives hang in the balance.


Vaccine Primer
Vaccines prepare the body against viruses by triggering an immune response. Traditional vaccines introduce a harmless piece of a virus to the body. Messenger RNA (mRNA) vaccines introduce a piece of mRNA that corresponds to a piece of a protein. With this mRNA blueprint, cells produce the viral protein. In both RNA and DNA of vaccines, the body activates a normal immune response, producing specialized proteins called antibodies that help the body protect against infection. Antibodies recognize viral particles, attach to them, and mark them for destruction. The antibodies remain so that the immune system can quickly respond in the future if the body sees the virus.


Complementary Data

Toward the end of his postdoc at Scripps, and shortly after he started at UW, Kirchdoerfer took a revolutionary step in the field of structural biology. He used cryo-EM to image a protein vital to the coronavirus replication process.

After SARS-CoV-2 enters a host cell, viral RNA is translated by the host. Then, virus polymerases (replicative enzymes produced by the host cell) generate new RNA genomes and messenger RNA. These RNAs, in turn, create the components needed to assemble new virus particles. Though much remains unknown, scientists do know that the 16 nonstructural proteins in SARS-CoV-2 — including nsp12, nsp7, and nsp8 — are critically important to these processes. By imaging the nonstructural proteins (named as such because these proteins are produced by the virus but aren’t components of a virus particle), scientists can learn more about RNA synthesis, processing, and replication. This knowledge can then contribute to new antivirals that halt virus replication and lessen the severity of infection.

Kirchdoerfer was the first to image the nsp12-nsp7-nsp8 complex from any coronavirus. When SARS-CoV-2 emerged, he wanted to image the nsp7 and nsp8 from that virus individually, but with cryo-EM, these subunits are too small to study on their own.

For this project, Kirchdoerfer turned to a different imaging method — and the people using it at UW. At the top of his list were biochemistry professors Katherine Henzler-Wildman and Chad Rienstra, codirectors of the National Magnetic Resonance Facility at Madison (NMRFAM), a campus-wide and national facility housed in the biochemistry department.

Biochemistry professor Katherine Henzler-Wildman stands on a platform attached to Fleckvieh 1, a Bruker Avance III HD 900 MHz instrument, which she used to image nonstrucutral coronavirus proteins, at the National Magnetic Resonance Facility at Madison. Photo by Michael P. King

“We were talking to Rob Kirchdoerfer, and we asked what we could do that would be helpful. He said that we could look at things that are too small for cryo-EM or things that we want to look at in solution rather than in crystals,” says Henzler-Wildman, who is the Jean V. Thomas Professor in Biochemistry.

Nuclear magnetic resonance (NMR) spectroscopy would provide insights into the SARS-CoV-2 nonstructural proteins, the team decided.

Over the next year, the group would tackle nsp7, nsp8, and a structural protein called the membrane protein. Their first experiments, performed using proteins produced by Kirchdoerfer’s lab and as part of an international consortium called the COVID19-NMR Project, confirmed that SARS-CoV-2 nsp7 and the nsp7 in the original SARS virus are similar both in their NMR signals and in their actual structure.

The scientists didn’t find this surprising — depending on which protein in the RNA replication machinery you compare, 94–96% of SARS and SARS-CoV-2 amino acids match. In an ideal world, the scientists’ studies on nsp8 and the membrane protein (which has 91% sequence identity with the membrane protein in the original SARS) would be just as straightforward. But that isn’t how research often progresses.

“We can’t say much more right now, but our results for nsp8 aren’t what we expected,” Henzler-Wildman says. She and her collaborators thought nsp8 would be flexible and would have multiple different shapes. What they found was that its structure depends on how much nsp8 is in solution. Now, the scientists need to figure out what nsp8 does in solution as well as how it ends up in its various configurations.

The NMRFAM team faces different challenges with the membrane protein. Perhaps counterintuitively, this protein doesn’t want to fold into a well-behaved three-dimensional structure, so scientists haven’t been able to produce it in the lab. So, while virologists believe the membrane protein plays an important role in how coronaviruses exit host cells, it remains understudied — and, some experts say, underutilized — in the fight against COVID-19.

By fall 2021, the UW–Madison team was making significant progress at purifying the membrane protein and was contemplating their next steps. Data they collect may be especially critical — research conducted a few hundred yards away suggests that this protein might be an active ingredient in the next COVID-19 vaccine.


NMR 101
While cryo-EM can handle imaging larger complexes at near-atomic resolution, NMR is restricted to small proteins. In NMR experiments, biological samples are placed in a probe that sits inside a powerful magnet. The probe contains a special type of antenna called a coil, which communicates with atomic nuclei in the sample. Radiofrequency signals are sent to the sample through the probe in a series of magnetic pulses that cause the nuclei in the sample to respond. Information from these interactions is picked up by the probe and sent to computers for analysis.


A Pan-Coronavirus Alternative?

When the pandemic hit, preeminent virologist Ann Palmenberg, known for sequencing the genetic code of the common cold virus, had been identifying molecular interactions between rhinovirus C (a virus closely linked to wheezing and asthma) and its cellular receptor.

Palmenberg’s team was using a chip technology created by a group of UW–Madison scientists, including biochemistry professor Michael Sussman. Each chip contains the entire genome of a virus in the form of protein fragments. By identifying where antibodies stick on these fragments and then comparing this information to viral structures obtained with cryo-EM, scientists can improve our understanding of immune responses to viruses.

“We were just about to make the next batch of chips and collect data [on rhinoviruses, which cause the common cold] . . . when COVID-19 came,” recalls Palmenberg, a biochemistry professor and Institute for Molecular Virology affiliate. “We said, you know what, instead of designing the rhinovirus sequences on this chip, let’s put coronavirus sequences on it.”

Palmenberg and her collaborators at the UW School of Medicine and Public Health decided to pivot their long-standing rhinovirus C project to study how protein snippets from SARS-CoV-2 and the six other coronaviruses known to infect humans responded to plasma samples from two groups of people — patients with COVID-19 and individuals who hadn’t been exposed to the virus. Nimble Therapeutics — a Madison-based company with ties to the biochemistry department that was spun out of Roche Sequencing Solutions — played a critical role by building the chips at a substantial discount.

Viroloist Ann Palmenberg redirected a rhinovirus C research project so she could lend her expertise to a study of SARS-CoV-2 on the UW campus. Photo by John Maniaci/UW Health

Kirchdoerfer, also an Institute for Molecular Virology affiliate, helped match antibody-sequence pairs from the protein chips to cryo-EM structures.

Results of the study demonstrate that humans mount strong, broad antibody responses to the spike and membrane proteins along with a third type of structural protein, the nucleocapsid.

“The signal from the membrane protein, in fact, is about six times stronger than from the spike protein, and future iterations of vaccines will be able to take that into account,” Palmenberg says.

The scientists suggest that membrane proteins could be a promising target for future diagnostics, vaccines, and therapeutics for several coronaviruses. They note that the immunogenicity of spike-based mRNA vaccines is variable, and not all individuals who get COVID-19 produce detectable antibodies against the spike or nucleocapsid proteins.

“There’s a lot of doors that were opened through this research,” says Palmenberg. “It’s possible that the cross-reactivity that can be conferred across coronaviruses is not the spike but the membrane protein. That’s where research is going to go . . . If you have immunity against a spike or a membrane protein, does that confer immunity against every other kind of coronavirus?”

Signal from Noise

Although his role might not always be obvious, Kirchdoerfer has contributed to many such projects that aim to understand SARS-CoV-2. He’s played an important role, for example, in devising new strategies to characterize the activity of enzymes, which catalyze biological processes. This project, led by Sussman, is expected to be central to rapid characterization of enzymatic activity in SARS-CoV-2.

Kirchdoerfer and his students, along with Henzler-Wildman and Rienstra, have contributed to a large, multi-institution project that standardized the production of SARS-CoV-2 proteins for screening and structural biology applications. And throughout the pandemic, Kirchdoerfer’s lab has been manufacturing and shipping high-quality proteins to labs around the world to assist with SARS-CoV-2 research.

Yet, cryo-EM remains at the core of Kirchdoerfer’s work.

“Rob works on a number of fronts surrounding how coronaviruses function, from isolated components to intact viruses,” says Elizabeth Wright, a biochemistry professor and affiliate at the Morgridge Institute for Research. “To investigate how SARS-CoV-2 replicates, he has started by assembling and examining the structure and function of components of the virus replication complex.”

Wright directs the Cryo-Electron Microscopy Research Center, which provides services to UW investigators who are working on SARS-CoV-2 and other projects. “Rob does the fundamental molecular biology, protein expression, and functional assays in his lab to determine if samples are of sufficient quality for cryo-EM imaging,” Wright says. “We then support him during the sample preparation, imaging, and initial data processing steps.”

Kirchdoerfer’s work on RNA polymerases represents a turning point in structural biology. “Our structures opened the door for coronavirus polymerase structural biology, and I think now there’s 25 or 30 structures of coronavirus polymerases in the database, all from the last two years,” he says. “We were really the first ones, I’d say, to kick that door down.”

His ongoing studies on coronavirus replication and transcription may be especially important as SARS-CoV-2 evolves. Virus function is often coupled to structure, and that’s to our advantage as scientists combat COVID-19 on all fronts.

“A virus has to be able to bind and interact with host cells and undergo fusion, so the structural elements in the spike protein [that accomplish those functions] are going to be conserved. When a virus mutates, it has positions that can also mutate, and that’s how you can get antibody escape or receptor switching and that sort of thing,” Kirchdoerfer says.

But understanding the SARS-CoV-2 replication machinery and characterizing the prolines inserted into proteins for vaccine candidates isn’t his end goal — it’s just the beginning.

“During an outbreak, there’s intense scientific interest, but as soon as that outbreak ends, interest also ebbs,” he says. “What I would like to do with my lab is more pandemic preparedness and looking for the next virus that’s going to cause a pandemic.”

To that end, Kirchdoerfer is studying other coronaviruses and other virus families, and he’s launching projects on viral entry in collaboration with classical virologists, veterinary medical biologists, and epidemiologists at UW. After all, as Kirchdoerfer says, the best way to battle a novel virus is to already understand how viruses work.


Sidebar: Coronavirus Education in Classrooms and Communities

Formally and informally, in lecture halls, by email, and even on city buses, biochemistry faculty advance greater understanding of COVID-19. Read more


This article was posted in Basic Science, Features, Health and Wellness, Spring 2022 and tagged , , , , , , , , , , , , , .