Deep mutational scanning of the Nipah virus fusion protein F using pseudoviruses maps ~8,500 single-residue effects, showing F is highly constrained and identifying antibody-escape mutations.
0:00Welcome to Base by Base, the papercast that brings genomics to you, whatever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Today involves a bit of a nightmare scenario.
0:10But I promise, the science behind it is actually incredibly hopeful. I like the sound of Hopeful, but you said nightmare scenario, so I'm going to need you to elaborate on that first. Okay, so I want you to imagine a killer.
0:22A pathogen. It's not new, but it is. Well, it's terrifying. It's case fatality rates. It's somewhere between 40% and 90%. Whoa, hold on, 90%. That's that's catastrophic. I mean, we panic about flea strains that have a one or 2% fatality rate.
0:3990% is basically a death sentence. It is, for context, that is significantly higher than almost any respiratory virus we deal with. And the way this thing operates, it's like a shapeshifter. Meaning what?
0:49It changes its disguise or something. Yeah. It relies on a mechanical transformation. It has a key to unlock your cells. But that's not really the most dangerous part. The real problem is the battering ram.
1:00It carries this fusion protein that acts like a loaded spring. It's, you know, ready to fire. Okay. And once it finds a target, it snaps irreversibly in milliseconds. It physically forces the virus and your cell to merge into one.
1:15That mechanical action, that snap is what lets the infection in. Exactly. So here's the big question for today. How do you design a defense against a biological machine that snaps shut faster than you can blink and even crazier?
1:28What if we could map out every single possible way this machine could break or evolve before it even happens? That sounds like the holy grail of pandemic preparedness. Instead of reacting to a virus after it mutates, you predict it's every move beforehand.
1:41That is exactly what we're talking about today. Okay, I'm Hook. are we looking at? We are looking at this tour de force study that essentially tried to map the future of a deadly virus. And we really have to give a huge amount of credit to the research team at the Fred Hutchinson Cancer Center and the University of Washington.
1:59Right. Specifically, the lead researcher, Brendan B. Larson, and Jesse D. Bloom. I was looking at the data, and the scale of what they attempted is just, it's pre-mind blowing. It really is. Their whole mission was to push our understanding of the Nepavirus fusion protein.
2:16And they did it using a technique called deep mutational scanning. Okay, so let's back up and set the stage. You mentioned Nepavirus. I feel like I hear about this one in the news every now and then, usually in a don't panic, but keep an eye on this kind of way.
2:29It's definitely a watch closely virus. Nipa is a batborn paramixo virus. It was 1st identified during a really big outbreak in Malaysia back in 1998. Right. That outbreak was just devastating for the pig farming industry there, and tragically for the farmers as well.
2:43And it hasn't just gone away since then, has it? No, not at all. We see these recurrent spillovers. So it jumps from nature into humans, mostly in Bangladesh and India. And when you say spillivers, that implies it's coming from animals.
2:56Usually bats, sometimes bats to humans, sometimes bats to another animal, then to humans. For instance, drinking raw date palm sap that bats have contaminated is a known root of infection. And the result is bad.
3:09It's severe. It causes fatal encephalitis swelling of the brain. And the real gap here. The reason this research is so vital is that we have no approved vaccines or therapeutics for it. Nothing after all this time.
3:22Nothing approved yet, which brings us right back to that battering ram. To stop Nipa, we have to understand the mechanics of how it breaches our defenses. Right. So let's get into the mechanics. You said the virus has two main tools on its surface.
3:35Correct. Two glycoproteins. Think of them as the virus' hands. First, you've got the receptor binding protein or RBP. That the key. Its job is to find the lock on your cell receptors called FNB2 or B3.
3:48Okay, so RBP grabs the doorknob, but just grabbing the doorknob doesn't open the door. Precisely. RBP just holds on. The violence is committed by the 2nd tool. The F protein, the fusion protein. This is our battering ram.
4:01And this F protein is the subject of this deep dive. It is. The F protein is what's called a class I fusion protein. To stick with the analogy, think of it like a mousetrap that's been said. It's in this betastable profusion state, just full of tension, potential energy, waiting for a trigger.
4:16So RBP finds the cell, turns the key. And that triggers the F protein. The spring releases. It shoots out a fusion peptide, like a harpoon that physically stabs into the wholesale membrane. Then the whole protein collapses into a new shape, the post-fusion state.
4:31And that action pulls the virus and the cell together until they merge. And once that spring snaps, you can't reset it. Nope, it's a one-way trip. And understanding this mechanism is just, it's absolutely critical. For a vaccine, you have to train the immune system to recognize that cock spring shape.
4:48If your antibodies only recognize the snapped post-fusion shape, it's too late. The virus is already in. So we need a map of the spring. But here's the problem, and I'm guessing this is why this paper is such a big deal.
4:59You can't just brew up a ton of NEPA virus in a standard lab and start tweaking it, right? That sounds, well, reckless. Oh, it would be incredibly dangerous. Nipa is a bio safety level 4 pathogen. That's the highest level there is.
5:13We're talking space suits, airlocks, the whole deal. Right. You cannot safely test 8000 different mutations of live NIPA. It's just not feasible. The risk would be astronomical. So how did they pull this off?
5:26They used a brilliant workaround? Pseudoviruses? Fake viruses. Kind of. Think of them as mannequins or stunt doubles. These are safe lentiviral particles, basically harmless shells. They're dressed up to look like Nepo on the outside.
5:39They carry the Nipa F protein. But they don't have the genetic code to replicate. So they can get into a cell, but they can't throw a party once they're inside. Exactly. They are one hit wonders, get in once, and that's it.
5:50This allowed the team to use a technique called deep mutational scanning. I love that term. Sounds so high tech. But what does that actually mean? Are they like scanning it with a laser? Huh, no, it's biology at a massive scale.
6:01It's brute force science, but done very elegantly. They synthesize what are called illegal pools. Oh, ego pools. break that down for me. So an olega nucleotide is a short strand of DNA. They basically printed 1000s of tiny snippets of DNA, and each one coded for a slightly different version of the F protein, a different mutation.
6:20They built a whole library. Wait, how big a library? How many mutations are we talking? They made 8,449 different mutations. That's 98.2% of every possible single amino acid change in the entire protein.
6:32They systematically broke the machine in almost every way you can imagine. That is just wild. It's like taking a car engine and one by one, replacing every single screw with a different one just to see which one makes the engine explode.
6:45That's a perfect analogy. And to test if the engine still ran, they use these stunt double viruses on cells in the lab. But not just any cells. What do they use? They use CHO cells, Chinese hamster ovary cells.
6:58But they engineered them to express bat receptors. Specifically from the Black Flying Fox, Terapis Electa. Huh. Why use bat receptors? Why not human ones? Two very smart reasons. One is biological relevance.
7:11Bats are the natural reservoir, so this is the virus's home turf. But the second, and this is key, is safety. By using bat receptors, they avoid any risk of accidentally teaching the F protein how to be better at infecting human cells during the experiment.
7:26Oh, that's a clever layer of safety. I hadn't even thought of. So they have their safe viruses, 1000s of mutations. What are they actually measuring? Two main things. Function and antigenicity. Function is.
7:36Does this mutation break the machine? Can the virus still get in? And antigenicity is, if the virus has this mutation, Can our antibodies still stop it? Let's get to the results then. Where is the armor weak?
7:47Finding number one is the F protein fragile. Surprisingly, yes. The data show the F protein is very stiff, or what we'd call highly constrained. It doesn't tolerate change well. The team had studied the RVP before, the key protein, and that was pretty flexible.
8:03It can handle a lot of mutations and still work. But not the F protein. Not at all. Most mutations just broke it. It stopped working completely. So it's a diva. It has to be perfect or it just doesn't perform.
8:14Exactly. And that lines up perfectly with what we see in nature. When you sequence NIPA from different outbreaks over the years, the F protein is 99% identical. It evolves very, very slowly. because it can't evolve quickly without basically self-destructing.
8:30That's good news for us, right? A static target is way easier to hit. It's fantastic news. But the study went deeper. They mapped the geography of vulnerability. They looked at the 3D shape of the protein to see where it can handle change.
8:44And where are the weak spots? The apex, to the very top of the protein is relatively tolerant. It can handle some mutations, but the sides, the lateral face, and the core are highly constrained. And the most protected part of all, the fusion peptide.
8:59Harpoon. The harpoon. It's tucked away in a little groove on a neighbor protein. The data show that if you change that groove even slightly, the whole machine fails. The harpoon can't launch properly. You mentioned a neighbor protein.
9:13The F protein doesn't work alone. No, it's a team sport. They group together in what we call trimers, groups of three, but this data suggests something even more complex. It looks like these trimers might pack together into hexamers, so groups of six.
9:26A 6 pack of battering rams. And the points where they touch are absolutely crucial. The study found that if you mutate those interfaces, you break the virus. Okay, let's talk vaccines. You said earlier, we need to catch the spring before it snaps.
9:40Did this study help with that? Significantly. To make a good vaccine, you need to show the immune system that prefusion shape. But proteins can be a bit wiggly, in a vaccine vial, they might snap on their own.
9:53If they do that, they're useless. You need to lock it in place. You need a safety pin. And the study found one. They identified specific spots where if you swap in an amino acid called a proline, it acts like a lock.
10:04Why, a proline? What's special about it? Prolines are really rigid. They put a kink in the protein chain. So by putting them in these key spots, you essentially freeze the protein so it can't extend into that post-fusion shape.
10:17It stabilizes the whole thing, making it a much, much better vaccine candidate. That is just, it's incredibly practical science. But what about treatments? For when someone's already infected? We need antibodies.
10:29And viruses are famous for mutating to escape them. Right, and this is where the study gets really cool. They looked at a concept called antibody resilience. They tested 6 different monoclonal antibodies against their whole mutation library.
10:42Before you tell us the winner. What's resilience? How is that different from just being strong? Great question. Potency is how well the drug kills the virus. Resilience is how hard it is for the virus to escape the drug.
10:55You can have a super potent antibody, but if the virus only needs to change one tiny thing to make it useless, that antibody isn't resilient. It's fragile. I see. So you want an antibbody that corners the virus, gives it no moves left.
11:07Exactly. And in this study, one antibody was the clear star. One F2. What F2? What made it so good? It was just incredibly resilient. Even when the virus mutated at the exact spot where one of 2 binds.
11:21The antibody could still neutralize it to some degree. It was so much harder to escape. Why? Is it targeting something the virus just can't change? That's the idea. It's likely targeting a spot that is so critical for the mechanical snap that if the virus changes it enough to escape the antibody, it also breaks its own machine.
11:39It kills itself in the process. That's the sweet spot you want in a drug. Now, finding number five. This one felt like magic to me, predicting the future. It really does. This validates the whole lookup table idea.
11:51So Nipa has a cousin called Hendra virus. They're about 88% identical. Close relatives but still different. Right. So the team looked at their Niba map and said, okay, based on our data, antibody 4H3 should not work on Hendra.
12:05Because Hendra already has a natural mutation at a key spot. Exactly. At site 70, Nipa has a glutamine, but Hendra has a lycine. The map basically screamed, that's an escape mutation, and sure enough, when they tested it, 4H3 was useless against Handra.
12:19They predicted it before they even ran the experiment. Correct. And they did it again, predicting another anabottle, 189, would be weaker against Hendra, because of another difference. And they were right again.
12:31This proves the map isn't just for NEPA. It works for its relatives. That's a perfect segue to the implications. What does this all mean for the big picture? Well, first, for vaccine development, We now have a literal list of candidate mutations, those prolines, to create a super stable vaccine antigen.
12:50That could save years of trial and error. And for drugs. It changes how we pick our champions. When we design antibody cocktails, We shouldn't just look for what's most potent. We have to look for resilience.
13:01We need antibodies like one F2 that the virus just can't get away from. It's better to have a slightly less potent drug that always works than a super drug that can fail with one tiny change. Exactly. And finally, pandemic preparedness.
13:15This study provides a lookup table. Explain that concept again for us. Okay, imagine a new outbreak starts tomorrow. We sequence the virus and see a new mutation at, say, residue 200. Instead of waiting weeks to grow that virus in a BSL 4 lab to test it, scientists can just look at this data.
13:33They check residue 200 on the map. If the map says that mutation breaks the protein, we can relax a bit. If it says that mutation lets it escape our best drugs, we know immediately that we're in trouble.
13:43That's speed. That's a huge advantage in a race against a virus. But we have to be fair. What are the limitations here? Well, it's a pseudovirus. It's a model. It doesn't face replication pressure like a real virus does inside a body.
13:57And there was a technical thing they had to do. They had to cut off the protein's tail. The part that's inside the cell. Yeah, the cytoplasmic tail. That tail might be involved in some signaling or stability that this study couldn't see because it was missing.
14:09So you eventually have to confirm these findings with the real live virus. But as a roadmap, it's a massive leap forward. Absolutely, a game changer. So if we boil this all down to one key insight. What is it?
14:20The insight is that the Nepa fusion protein is a highly constrained mechanical device. Because it can't easily change its shape because braking, it's the Achilles heel of the virus. A static target that's perfect for vaccines at broad spectrum antibodies.
14:36And unlike the flu or HIV, which shift constantly. This thing is kind of stuck in its ways. It is, and that gives us the upper hand. I love that. Here's a thought to leave you with, though. We talked about Nipa and its cousin Hendra, but we know there are others out there.
14:51You hear about Longavirus, another henipavirus that popped up recently. Right. So the question is, what does this mean for our ability to predict the behavior of those cousin viruses, the ones we haven't even found yet?
15:03Can we use this NIPA map to build defenses for viruses that haven't even made the jump to humans? Can we be ready for the cousin before it even knocks on the door? That is the frontier. And with data like this, I think the answer is getting much closer to yes.
15:18Incredible work. Huge thanks to the team at Fred Hutch and UW for this. Indeed, it's fantastic. This episode was based on an open access article under the CCBY 4.0 license. You can find a direct link to the paper and the license in our episode description.
15:33If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating. If you'd like to support our work, use the donation link in the description. Now stay with us for an original track, created especially for this episode, and inspired by the article you've just heard about.
15:47Thanks for listening and join us next time as we explore more science base by base.
