Issue Date: September 7, 2015
Slowing Prions With Small Molecules
When people with prion diseases start to experience symptoms, such as memory loss and difficulty moving, they don’t have much time left. Most patients die in about a year.
Prion diseases start when proteins in the brain go rogue. For most people, neurons churn out normal, healthy prion proteins. But in about one in a million people, something triggers these proteins to misfold into a structure that then coaxes other prions to do the same. This process propagates through the brain, and misfolded prions eventually aggregate into fibrils that can destroy neurons.
In about 10–20% of patients, a mutation in the prion protein gene starts this wave of protein misfolding. For a tiny minority, the propagation process begins when a patient gets infected with misfolded prions from other people or animals, such as when people consume beef from cattle that had bovine spongiform encephalopathy, also known as mad cow disease. But for most people with prion disease, the exact trigger is unclear.
As is the case with more common neurodegenerative diseases such as Alzheimer’s and Parkinson’s, there are currently no therapies that can stop the progression of prion diseases. But in the past decade, a few research groups have reported compounds that disrupt prion propagation in infected mice and extend the animals’ life spans. The molecules don’t cure the condition, and none of these specific compounds are likely to make it into the clinic to treat human prion diseases, but the scientists who developed them believe the work offers hope that one day small molecules could serve as antiprion therapies.
That wasn’t always a given, says Armin Giese, a prion researcher at Ludwig Maximilian University of Munich. When Giese started thinking about developing antiprion drugs in the early 2000s, pharmaceutical scientists said they didn’t have a chance of working. Compounds that disrupt prion aggregation would have to target protein-protein interactions, and that is hard to do with small molecules, they told him.
“Small molecules are fine to fill a pocket in an enzyme or receptor,” Giese says, recalling the prevailing wisdom at the time. “But protein-protein interactions involve surface areas that are so huge that a small molecule just won’t do the trick.”
But in 2007, a Japanese research team reported the first orally available small molecule, called compound B, to delay death in prion-infected animals (J. Virol. , DOI: 10.1128/jvi.01563-07). Then in 2013, two separate research groups reported others they had found through high-throughput screens of compound libraries. The molecular mechanism behind these and other antiprion compounds is still not clear.
Giese’s group made one of the two 2013 findings. He and his colleagues ran screens with a few techniques, including a fluorescence-based method, to detect protein aggregation. They zeroed in on the compound anle138b, which more than doubled the life spans of mice infected with prions (Acta. Neuropathol. 2013, DOI: 10.1007/s00401-013-1114-9). Other groups are now testing the compound in mice with prion protein mutations that cause disease in humans (see box on page 38).
The other compound reported in 2013 was IND24 (Proc. Natl. Acad. Sci. USA, DOI: 10.1073/pnas.1317164110). Stanley B. Prusiner and Kurt Giles at the University of California, San Francisco, used a cell-based screen to find it and other 2-aminothiazoles that block prion aggregation. Prusiner won the 1997 Nobel Prize in Physiology or Medicine for discovering prions.
But not all of the IND24 findings were good news. The UCSF team found that IND24 was not effective when given to mice infected with a certain type of prion—one taken from the brains of mice that died after having been treated with the compound.
“It’s an interesting finding but a bit of a downer as well,” Giles says. This finding suggests that the compound spurs the development of resistant strains of prions in the brains of treated animals. When researchers talk about prion strains, they’re referring to forms of the proteins—probably arising from certain folded structures—that have a similar set of biochemical properties, such as stability against unfolding or rate of aggregation (see page 11).
IND24 also demonstrated a limitation seen with compound B and anle138b, says Neil Cashman, a neurologist who studies prion therapeutics at the University of British Columbia. These compounds aren’t effective against all prion strains. For example, compound B worked against a mouse prion strain but not a hamster one. And no antiprion compound reported to date has demonstrated the ability to slow the progression of human prion strains that have been injected into animals.
Giles isn’t surprised. The screens that produced anle138b and IND24 used cells infected with mouse prion strains, not human ones, to pinpoint the compounds. To find small-molecule therapies for human prion diseases, researchers probably need to run screens with cells infected with human strains, and these don’t exist.
“People have been trying to do that for 30 years, and no one has succeeded yet,” Giles says. The problem is that scientists don’t understand what makes one type of cell line able to propagate misfolded prions and another incapable of it. Out of the hundreds of cell lines researchers have tried to infect with prions, Giles points out, only about a half-dozen worked.
To avoid this and other limitations of cell-based compound screens, Adriano Aguzzi at the University of Zurich and coworkers recently took a rational, structure-based approach when designing their antiprion compounds. The researchers had been studying a class of compounds called polythiophenes that fluoresce when they bind to prion aggregates. This property makes them good for staining prions in infected brain tissue. Preliminary studies in mice suggested that the polythiophenes also had antiprion potential.
Aguzzi wanted to understand the structural rules for designing an effective antiprion polythiophene. Unfortunately, scientists haven’t been able to get atomic-level structural data on mammalian prion aggregates. So Aguzzi and his team turned to a fungal protein called HET-s that had been studied structurally. Although not homologous to mammalian prions, HET-s aggregates like them.
The team used solid-state nuclear magnetic resonance spectroscopy to determine the structures of polythiophenes bound to HET-s aggregates. Then the researchers used those structural data to run molecular dynamics simulations of the binding of different polythiophenes to prion aggregates. When they synthesized the modeled compounds and tested them in prion-infected mice, the researchers found that the life span expansions they observed in the rodents correlated with the polythiophene-prion binding strength calculated by the simulations, confirming their model.
Last month, Aguzzi’s team reported their best performer, LIN5044, and the structural lessons they learned (Sci. Transl. Med. 2015, DOI: 10.1126/scitranslmed.aab1923). The compounds needed at least five thiophene rings and had to sport carboxylic acid side chains about 10 Å apart—one on every other ring in the chain. In the prion simulations, these negatively charged groups helped the compounds anchor themselves into the grooves of the prion aggregates by interacting with positively charged lysine amino acids.
The structural insights gained from the study are more important than the specific compounds, Aguzzi says. “The molecules that eventually enter the clinic might not even have a polythiophene backbone,” he says. “Maybe you can do it with a completely different scaffold.”
Cashman agrees: “The breakthrough here is defining a rational target that can be exploited to develop potent compounds.” High-throughput screening methods are blind to targets on the prion aggregates, so they can yield compounds that don’t hit conserved sites across prion strains. By using a distantly related protein in his models, Aguzzi may have found a conserved target, and this could make the resulting molecules effective against multiple prion strains, including human ones, Cashman says.
Looking at all the recent work on antiprion small molecules, Aguzzi, Giles, and Cashman are optimistic that someone will find a compound that works against human prion diseases. “The positive message from our and others’ studies is that it’s possible to design compounds that work in animals,” Giles says. “There’s no voodoo here. You just need to systematically do the work, or you get lucky.” ◾
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society