NIAID Now | August 23, 2021
Studying infectious diseases in a research laboratory can be like a family building a jigsaw puzzle at home, though with an infinite number of pieces. Dozens of people will enter and exit the research team over many years, making important contributions to the puzzle but never finding the final piece.
In January 2021, Allison Kraus, Ph.D., found the final piece to learning the molecular structure of an infectious prion—a puzzle that her mentor, Byron Caughey, Ph.D., had started 30 years earlier. Together, they and a broad team of scientists at NIAID’s Rocky Mountain Laboratories (RML) in Montana and Case Western Reserve University (CWRU) in Ohio solved the first high-resolution, three-dimensional prion structure. They also obtained unprecedented, but lower resolution, images of another distinct prion strain. Determining prion structures, and their diversity, is fundamental in helping researchers to understand how prion diseases develop, how treatments could be targeted to slow and prevent disease, and how prions compare to proteins that cause related diseases like Alzheimer’s and Parkinson’s.
A Basis for Developing Therapeutics
Prion diseases are caused by prions, which are corrupted forms of a mammalian protein called prion protein, or PrP. Caughey thinks their study, “High-resolution structure and strain comparison of infectious mammalian prions,” published Aug. 23 in Molecular Cell, will provide colleagues with an initial example of how normal PrP molecules can be refolded and accumulated to cause disease.
“Based on the high-resolution structure that we can see, we think that pathogenic rod-like PrP fibrils have sticky ends that capture normal PrP molecules and unravel them as if they were binding to a narrow serpentine strip of Velcro,” he said.
Previously, scientists only had educated guesses about the structure of prions to guide their work. The two leading theories centered on the structure either being a series of stacked corkscrew-like molecules—the beta-solenoid model—or what Caughey’s group proposed in 2014 and now has proven correct: the PIRIBS model. PIRIBS, which stands for parallel‐in‐register intermolecular beta‐sheet, shows pancake-like layers of PrP molecules forming prion fibrils. Beta sheets are a key ladder-like substructure of such fibrils.
By comparing the new images of two rodent prion strains, Kraus and Caughey could see that they differ markedly in structure in ways that must account in part for the different neurodegenerative diseases that they cause. They acknowledge these initial structures will not explain all prion strains, and that they need to, as Caughey says, “humanize the understanding of prion structures.”
Their high-resolution structure is of a specific hamster-adapted strain of scrapie, a prion disease that occurs naturally in sheep and goats. The second strain that they have analyzed is a mouse-adapted form of scrapie. Now they are analyzing the extent to which prion structures vary among other prion diseases. By pinning down the first detailed structure of a highly infectious prion, their work also frames thinking about factors that control the extent to which various pathologic protein aggregates, such as those of Alzheimer’s and Parkinson’s disease, might also be transmissible.
With a near-atomic prion structure in hand, they are more able to begin assessing the changes that take place in the conversion of normal PrP to an infectious prion form. That will allow scientists, they believe, to target steps where treatment might be effective at slowing or preventing the spread of disease.
“We and other groups are seeking drugs to stabilize normal PrP molecules and prevent them from refolding abnormally,” Caughey says. “Others are trying to develop vaccines that could be designed to recognize prions and either block replication or help the body eliminate them. Further work in this direction should be greatly facilitated by accurate knowledge of prion structures.”
Prion diseases are transmissible, untreatable, and fatal brain diseases. Discussing them can get confusing: All mammals have PrP molecules that presumably perform normal physiological functions as single units. For unknown reasons, PrP molecules can refold and organize into highly structured assemblies that cause infectious disease. These are called prions—aka an infectious form of PrP molecules.
Prions can reproduce in hosts a billion times or more during infection. Remarkably, they do so without having their own infection-specific genes because all PrP molecules, whether normal or pathogenic, are encoded by the same host gene. The difference between these forms of PrP is primarily a matter of folding rather than an amino acid sequence. This means that scientists cannot easily track prions like they can viruses and bacteria. All prion diseases are fatal: Caughey says 1 microgram—one millionth of a gram—can contain 100 million lethal doses.
In 1990 Caughey first saw evidence of the beta sheets that would eventually lead to their PIRIBS models of prion structure. He and colleagues, using a purified sample of scrapie taken from hamsters, were the first scientists to get a glimpse of this aberrant structure. But that was using an infrared spectrometer, which measures absorbance of infrared light and does not provide structural detail.
Scientists often use rodent-adapted scrapie strains as prototypes for related prion diseases of humans (Creutzfeldt-Jakob disease), cattle (mad cow disease), deer and elk (chronic wasting disease). The researchers published their study of the beta sheets in prions in 1991, and ever since Caughey and the many young scientists he has mentored have worked to improve on that work. The hamster 263K strain of scrapie used then is the same as the one used in their recent study.
Kraus arrived at RML in 2011 from the University of Alberta to begin a post-doctoral fellowship working for Caughey. She wanted to learn more about how and why prion proteins folded and could cause disease. Early in the fellowship he discussed the prion structure project with Kraus. The work would rely on using a technique known as cryogenic electron microscopy, or cryo-EM to create images showing the structure of prions. RML is a central microscopy facility for NIAID and in 2009 its group received a Titan Krios EM, which could produce ultra- high-resolution images: imagine seeing the atoms that comprise a gold particle. At the time, RML’s Titan Krios was one of only four such microscopes in the United States.
Shortly after the Titan Krios arrived, Caughey and EM supervisor Beth Fischer began discussing how to apply the new technology to the prion structure project. They were confident the microscope could help scientists understand and model how the misshapen PrP molecules that cause scrapie are structured, replicate, and cause disease.
They happily added Kraus to the mix. For the next eight years she and other RML EM team members—Cindi Schwartz, Forrest Hoyt, and Bryan Hansen—spent countless hours together hypothesizing, testing, failing, reading, re-hypothesizing, re-testing, improving … over and over. While the goal was the pursuit of a high-resolution prion structure, the work along the way revealed many important novel insights into structural features of prions.
“The process was truly empirical science, in its grittiest, hard-earned way,” Kraus recalls.
A day in 2018 stands out among many key moments. “For the first time we were able to see mathematical evidence of the repeating units that form the building blocks of a prion fibril,” Kraus said. “No one had ever visualized them ... it was the first hard piece of evidence that, using cryo-EM, we could actually achieve the resolution we would need to solve the structure.”
In 2019 Kraus started her own independent laboratory as an assistant professor at CWRU, but importantly, the collaboration with Caughey’s lab at RML continued—and now Kraus had access to a newer, improved cryo-EM instrument. Utilizing protocols and a pipeline developed with the team over years, Schwartz would prepare samples to send Kraus for imaging at CWRU, and Kraus would send RML terabyte after terabyte of data for Hoyt and Hansen to analyze and incorporate into the developing model.
By fall 2020—during a pandemic—the groups produced images of scrapie prions that were unprecedented in resolution, and utilizing recently developed computational methods, showed immense promise to yield the first atomic level model. Knowing they were close, Kraus recalls working day and night through much of November and December 2020 and into 2021.
“There were many Zoom meetings, phone calls, texts and remote computing challenges” involving about 13,000 high-resolution images, Kraus says. “This study and the team members covered so many different aspects of science, from prion biology to prion biochemistry, microscope technology, software advances, mathematics, computer graphics to bring this all together.”
Then in early January 2021, the final piece: “When the massive data sets had been analyzed, and for the first time we realized we could visualize, at high resolution, the structure of a prion. It was such a fun Zoom call. We spent the afternoon exploring what we had hypothesized correctly, what we hadn’t, and what we hadn’t even thought about.”
“Finally, we got a clear view of its dastardly mugshot,” Caughey added.
And now, they know they’re not finished with the puzzle after all. Kraus and Caughey say that their first fully infectious prion structure is one of many that will be needed to understand the full range of these deadly, unusual, protein pathogens.
Or, as Kraus puts it, “the ‘final piece’ of this 30-year puzzle is but the first piece in a new project to target these aberrant protein structures.”
Reference: A Kraus et al. High-resolution structure and strain comparison of infectious mammalian prions. Molecular Cell. DOI: 10.1016/j.molcel.2021.08.011. (2021).