This study uses antisense oligonucleotides (ASOs) to lower prion protein (PrP) RNA in mice and shows dose-dependent extension of survival, efficacy across multiple prion strains, reversal of molecular biomarkers, and benefit even when treatment is delayed into symptomatic stages.
0:00Welcome to Base by Dace, the paper cast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. In August 2026, we publish 2 episodes dedicated to Toledo Sousa, the Brazilian Aviation Communicator behind the channel of Yonis E Muska's Airplanes and Music, who shared publicly that he had been diagnosed with Krozfelt Jacob disease.
0:21We said then that there were 2 episodes. We were wrong. We have decided to keep going. What began as a dedication has become a series on Kurzfelt Jacob disease, because the science itself has a sequence, and 2 episodes were not enough to follow it.
0:34Today is the 3rd and it is the missing link between the other two. Right. So, uh, to really frame where we're going today. I want you to think about the massive chasm between a promising lab result and a viable human drug.
0:48Yeah it's huge. Like, imagine you're trying to invent a master key, right? Okay. And in the lab, you manage to carve a key that perfectly opens one specific door under perfect temperature conditions on a Tuesday.
0:59And that's incredible. You know, it's a breakthrough. But you can't sell it to a 1000000 people with a 1000000 different doors and claim you have a master key. Right. You can't. So what does it actually take to forge a key that works in the, you know, the messy generalized reality of clinical medicine?
1:18Well, that distinction between a localized finding and a generalized drug, that's that's basically the core challenge of pharmacology. Because finding success in one disease strain, at one specific dose, at one exact moment in the disease progression, that is merely a finding.
1:36It's just a single data point to transform a finding into an actual medicine. A treatment has to survive this just brutal test of generalization. brutal test. Yeah, it has proved itself across different variations of the disease at different stages of illness, and you have to provide measurable proof that is actually reversing the underlying pathology.
1:56Not just masking it. Exactly, not just masking it. And that brings us exactly to the focus of today's deep dive. We're exploring the exact moment a specific finding survived that brutal test and became a generalized hypothesis.
2:09So to keep our chronological sequence crystal clear for everyone following the series, let's map out exactly where today's discussion sits. Let's do that. So in our 2nd deep dive of this series, we covered a 2019 study by Raymond and colleagues.
2:23Right, the 2019 paper. Yeah, and that was a milestone because it showed, for the 1st time in living animals, that lowering the prion protein extends survival even when you begin treatment after symptoms have appeared.
2:35Yes. And that 2019 study was the critical sort of one door on a Tuesday proof of concept. But then, in our very 1st deep dive of this series, We actually jumped way ahead in time. We did. To the 2026 Dalen Serranay paper.
2:49Yeah. And that 2026 paper featured a completely different molecule. Uh, totally different delivery mechanism. It represented a much later chapter in the scientific journey toward a cure. Right. So today we're filling in the gap.
3:02We are. Today's deep dive is the 2020 paper that sits right in the middle. It's the missing link. It is the bridge. And the reason this 2020 paper is so critical. is that a single successful experiment does not start a clinical program.
3:16No. Generalization does. So this 2020 study took the same approach from 2019, and they tested it across 5 different pryon strains. Wow. Of course, a whole range of doses, at multiple different stages of the disease, and against several independent measures of whether the animal was actually getting better.
3:34This is the exact moment, a mere result, solidified into a drug hypothesis, robust enough to take into people. That's amazing. But before we look at how they built this, you know, ultimate scientific obstacle course.
3:47We need to acknowledge who built it. Because the collaboration behind this work is unique. It is a really remarkable partnership. Yeah. This 2020 study was driven by a collaboration between a team at the Broad Institute, led by Sonya Velab and Eric Medical, and Iona's Pharmaceuticals.
4:03And I want to specifically highlight, and I own a scientist named Hinzau, who is the 2nd author on this 2020 paper. Hmm. Thou also co-authored the 2019 study. providing a really crucial thread of scientific continuity.
4:17Right, connecting the two. Exactly, from that initial finding to this massive generalization effort. And, um, as a matter of public record, which she has spoken about openly, it's important to state factually that Sonya Velov carries a pathogenic variant in the prion protein gene.
4:33She is dedicating her life to developing treatments for the very disease she is at risk of developing. It grounds the science in absolute reality. Yeah. The stakes for getting this right. for proving that this therapy can generalize to human beings safely and effectively.
4:48They are absolute. Yeah, yeah, absolutely. So let's unpack the tools being used here and the background biology. We're talking about ASOs, uh, anti-sense, oligonucleotides, and their target is the prion protein or PRP.
5:01Yes. But I have to ask about the logic of the target itself. Like, can we really just target a native protein in the brain, basically shut off its production and expect the patient to be okay? That's great question.
5:12doesn't the human brain need that protein to function normally? I mean, it's there for a reason, right? So this raises one of the most fascinating aspects of prion disease. Okay. The prion protein. PRP is highly unusual in mammalian biology.
5:26It is almost entirely dispensable for normal health. Wait, really? Yeah. We know this definitively, because animals that have been genetically engineered to lack this protein entirely, so-called knockout mice, they're healthy.
5:41Wow. They develop normally. They live normal lifespans. So they just don't need it. Right. But while the healthy body doesn't seem to need PRP, the disease absolutely requires it to progress. Ah, I see.
5:53Brian disease is caused by the misfolding of this native protein. So if the native protein isn't there, the disease has no raw material to corrupt. Okay, so it's the perfect target. You're removing the fuel and the fire goes out.
6:05Exactly. But the house itself doesn't actually need that fuel to stand. That's great way to put it. So how do these ASOs actually remove the fuel, though? Like mechanistically what is happening inside the cell.
6:16Right, so ASOs are essentially short synthetic strings of genetic code. Okay. They're designed to match and bind perfectly to a specific messenger RNA. In this case, the RNA that carries the instructions to build the prion protein, known as PRNPRNA.
6:32Right. When the ASO binds to this target RNA inside the cell, it creates a hybrid molecule. It's part RNA part synthetic DNA. Oh, okay. Now, the cell's own internal machinery recognizes this hybrid structure as abnormal.
6:45Oh I see. It triggers a natural cellular defense mechanism and deploys an endogenous enzyme called RNA's H. So if I'm following the mechanism here, the ASO is basically acting like a laser targeting beacon placed on the RNA, right?
7:00Yes. And RA's H is the drone strike that sees the beacon and destroys the factory instructions. That is a highly accurate way to visualize it, actually. Okay, good. Arne's age acts like a pair of molecular scissors.
7:11It cuts and degrades the RNA that the ASO has tagged. Wow. Because the genetic instructions are destroyed before they can be read. The cell simply stops manufacturing the prion protein. The factory is shut down.
7:22The factory is shut down. That makes perfect sense for lowering the protein, but here is where we hit the historical roadblock. strains. Yes, the concept of prion strains. Because when I hear strain, I normally think of a virus mutating its DNA, like, you know, the flu.
7:37But prions don't have DNA or RNA, they are just proteins. Exactly. So what exactly is a pryon strain? And why have they defeated older drugs? Okay, so a prion strain is essentially a variation in the three-dimensional shape or conformation of the misfolded protein.
7:55Okay. Think of it like origami. Origami, okay. You can take the exact same sheet of paper, the exact same amino acid sequence and fold it into a crane, or a box, or a frog. Oh I see. These different toxic shapes are the different strains.
8:09And this has been the downfall of previous enterprise on drugs. Because they only target one shape. Exactly. Many older drugs were designed to bind to and neutralize one specific misfolded shape. But pryons are adaptable.
8:22If you use a drug that only blocks the crane shape, the prion can shift its confirmation, effectively mutating into the box shape. Oh man. Yeah, a drug resistant strain that the older drug can no longer recognize.
8:35Okay, let me let me deduce something here based on the mechanism we just discussed. Go for it. If the ASO is just shutting off the factory that produces the normal unfolded paper. The actual shape of the toxic origami shouldn't matter at all, right?
8:49That is the exact logic. By targeting the native protein, they are completely bypassing the prion's ability to mutate and escape. Yes, and that is the exact hypothesis they needed to prove. Right. But to prove that this approach truly solves the strain problem and is ready for the clinic, the researchers had to design an incredibly rigorous set of experiments.
9:09They couldn't just assume it worked, they had to prove it mathematically and biologically. Which brings us to the core methodology. And instead of just listing the tools, let's look at this as a problem solving exercise.
9:20If you're a scientist, how do you mathematically prove a mouse's brain is healing from a neurodegenerative disease without opening it up? Right. How did they set the baseline for this obstacle course? The rigor of this setup is really what separates a finding from a clinical candidate.
9:38First, they induce the disease in mice using an interest cerebral inoculation. Okay. They injected a one% brain homogenate directly into the brain. And just so we're completely clear for the listeners, a one% brain homogenate means a precisely measured blended mixture of infected tissue.
9:57Yes. They do this to ensure every single animal in the study gets the exact same dose of the disease at the exact same starting line, correct? Correct. It creates a highly standardized infection. Then, they delivered the ASO treatment using an interest or a broventricular or ICV bolus.
10:13Okay, an ICV bolus. Yes, this is a direct injection into the cerebrospinal fluid. Why not just an IV? Well, they use this route because ASOs are large molecules that do not easily cross the blood brain barrier from the bloodstream.
10:25Ah, I see. So injecting them into the fluid, bathing the brain. Allows the drug to distribute deeply into the brain tissue. Got it. And they tested this treatment using 2 very different timelines, right?
10:38prophylactic dosing and delayed dosing. Yes. prophylactic, meaning giving the drought early, well before clear symptoms or major pathology appear, and delayed dosing meeting, waiting until after the disease has established a real foothold in the brain.
10:53It's like, it's the difference between boarding up your windows a week before the hurricane hits versus trying to patch the roof while the wind is howling and it's already raining inside the house. That's a perfect analogy, and the FDA needs to know what happens in both scenarios.
11:08Right. And you can't just wait to see how long it takes for the mice to die to measure success. You need real-time trackers. You need multiple independent readouts to prove the animals getting better at a molecular, cellular, and behavioral level.
11:21So what were the specific readouts they used to track this in real time? Well, at the molecular level, they use QPCR. Which is a standard laboratory technique used to amplify and literally count the specific genetic material present, right?
11:34Exactly. So they used QPCR to measure the actual reduction of PMP RNA in the brain, proving the drug successfully hit its target. Yes. Then they needed to measure the actual damage to the brain. Okay. For this, they looked at biomarkers.
11:48They measured plasma, neurofilament light, or NFL. NFL. Yeah. NFL is a structural protein that acts like the scaffolding inside healthy neurons. When neurons are damaged or begin dying. They break open and spill this NFL protein into the bloodstream.
12:04Oh wow. So tracking plasma NFL gives you a direct real-time marker of brain injury. That solves the problem of tracking injury without opening the brain. It does. What about inflammation? To track inflammation, they use live imaging to measure something called GFAP bioluminescence.
12:20Okay. They utilize genetically modified reporter mice. These mice are engineered so that when they experience astrocytosis, which is a massive inflammatory response by the brain's immune cell. Their brain's actually emit light.
12:33Wait, they glow. They literally glow. The provider of the glow, the worse the inflammation. The mice literally glow based on how inflamed their brains are. That is an incredible tool. And they didn't stop at the molecular and cellular levels.
12:46They tested the mice behaviorally. They used the rotorod test, uh, placing the mice on a spinning cylinder to rigorously check their motor function and balance as they try to avoid falling off, and they tested their nest building abilities.
13:01Let's pause on the nest building for a second. Okay. Why give a mouse cotton squares and paper? What does a failing mess look like compared to a healthy one and why does it matter? So nest building is a highly complex instinctual behavior for a mouse.
13:16A healthy mouse will take those raw materials and construct a really complex three-dimensional enclosed structure for warmth and safety. It requires planning, coordination, and sustained attention. Makes sense. A mouse experiencing cognitive decline will simply leave the materials scattered across the cage floor, or perhaps just pile them flatly in a corner.
13:37Oh, that's sad. It is, but it serves as a highly reliable, observable proxy for their cognitive and behavioral health. Okay, so the stage is set. The hurricane is coming, the windows are boarded, or the roof is leaking, and we have the tools to measure every drop of rain.
13:53Yes. Let's get to the key findings, as stated in the paper. Let's start with the dose. I'm assuming you don't need to destroy 100% of the RNA to see a clinical benefit. Not at all. The results were strictly dose dependent.
14:07Okay. The degree of PNP RNA reduction directly tracked with how long the animal survived. But the critical detail for human translation is that even a modest knockdown, roughly a 21% reduction in the RNA, yielded a measurable significant survival benefit.
14:24You just have to slow the factory production by 20% to buy more time. And this benefit held up in both the prophylactic and the delayed setups. But if they if they give the drug early and keep the protein level suppressed, Does it mimic the genetic knockout mice we discussed earlier?
14:40It does. When they administered chronic early dosing, it extended survival so effectively that it matched the lifespan of mice that were genetically engineered from birth to carry only one working copy of the prion gene.
14:53Pharmacologically, they successfully recreated the protective effect of a genetic knockout. Now for the ultimate test, the strains. We know older drugs failed because pryans shifted their origami shapes.
15:06Did the data back up our deduction that shutting down the factory bypasses the strain problem entirely? It completely backed it up. The data demonstrated efficacy across 5 distinct prion strains. Five of them.
15:19Yeah, and to ensure they were testing a broad spectrum of biology, they tested RML 22L and ME 7, which are strains adapted from sheep and goat scrapy. Okay. They tested Fukuoko one, which is adapted from a human prion disease.
15:31Wow. And they even tested OSU, which is an entirely synthetic prion created in a lab. So they threw nature, human adaptation, and synthetic variations at it. Yes. And the ASO therapy extended survival against all five.
15:44That is incredible. It really is. And crucially, across all these diverse tests, absolutely no drug resistance strain emerged. None. None. Man, I want to pause on this next finding because this is the data point that actually allows a drug to enter human trials with confidence.
16:00What happened to the biomarkers, the NFL and the blood and the GFAP brain inflammation? When they gave the drug at 120 days post-infection, you know, after the disease had already taken hold. So 120 days post-infection, plasma NFL was already dramatically elevated, and the brain was heavily inflamed.
16:18When they administered a single dose of the ASO at this late stage. The trajectory didn't just pause. It reversed. Reversed. Wow. Yes. Plasma NFL level significantly dropped after the treatment. And the GFAP biodominescence, that glowing marker of neuroinflammation.
16:36It declined back down. Just to be clear, when we say it declined, we're comparing this to saline treated mice, right? Meaning mice that just received a placebo injection of salt water. The mice that got the actual drug, saw their markers of active injury move backward.
16:51Yes. Backward. This is what makes these measurements usable as endpoints in the clinical trial. Because if you can show a regulatory agency that your drug makes the NFL levels in a human drop. You have a quantifiable way to prove the drug is working without having to wait and see how long the patient survives.
17:11Exactly. And to our knowledge, this paper was the 1st time pharmacological reversal of a translatable biomarker of disease had ever been demonstrated in a pry on infected animal. Really? Yeah, it validates both the biomarker and the therapy simultaneously.
17:27We do need to look at the absolute limits of this drug, though. Right. There are limits. Because they push the delayed treatment to the extreme edge. They waited until the mice were in the late symptomatic stage.
17:37We're talking 132 to 143 days post-infection. Yeah, very late. At this point, the mice had already experienced clear observable physical symptoms. Severe weight loss, complete loss of nest building ability. Right.
17:51What happened when they intervened that late? Well, at those late symptomatic stages, a single dose still managed to prolong survival by months in a subset of the animals. However, the paper states plainly, while it extended their lives, it did not reverse the clinical symptoms themselves.
18:09They didn't regain the lost weight, and their nest building behavior didn't recover. By 156 days, when the disease in untreated mice was completely terminal, the therapy had no effect at all. There's a lag time for the ASO to take effect.
18:26And at the terminal stage, there simply isn't enough time left for the drug to clear the RNA. I want to push back on the biological definition of success here. Because if the drug extended their survival by months, but they didn't regain weight and couldn't build nests.
18:39Did the drug actually heal them? Or did it just prolong a state of severe biological decline? Yeah, that's the big question. Is that actually considered a clinical success if the underlying function isn't restored?
18:50It is a profound biological question. What it points to is that while PRP lowering is incredibly powerful at stopping the engine of the disease, the underlying neuromal damage eventually crosses a threshold where the clinical decline becomes locked in.
19:05Locked in. Yeah, the neurons are already gone. You can put out the fire, but you can't unburn the wood. It emphasizes why biomarkers like NFL are so vital. We have to catch and treat this disease before that point of no return is crossed.
19:20To summarize this incredible journey. We've seen how a localized finding from 2019 was systematically tested across 5 diverse prion strains across multiple doses and across both early and elite disease stages.
19:34Everything. They proved it wasn't a fluke. They proved it doesn't trigger resistance by starving the disease of its raw material, and they proved it can actually reverse translatable biomarkers of brain injury.
19:44They really did. This 2020 paper is the exact moment a robust, generalized hypothesis was born, ready to be taken into human trials. It serves as the blueprint for how you build a clinical program. You don't just prove that your drug works.
19:57You map exactly when it works, why it works, and how you're going to measure its success in a living patient using tools that translate from a mouse directly to a human. Which leaves us with a final thought to ponder.
20:09We just discussed that in those late stage symptomatic mice, lowering the prion protein extended their survival by months, even after physical symptoms were too far gone to reverse. It forces us to ask, where exactly is the biological point of no return in neurodegeneration?
20:26If we can stop the engine of the disease that late and still buy months of biological life, as delivery methods improve and we learn to intervene earlier, how much further back could we push that boundary?
20:37Could we one day stop the clock entirely? To all our learners out there, keep questioning, keep exploring, and will catch you on the next deep dive.
