EP 64 · 10:23

Restoring aspects of vision

From Nobel Prize 2026: The Science Gets Stranger

Episode
4/17
Watch Nobel Prize 2026: The Science Gets Stranger
In this chapter

A 2026 trial from a University of Pittsburgh-led consortium used optogenetics to partially restore vision in 10 people with advanced retinitis pigmentosa, a disease where light-detecting rods and cones degrade. Instead of targeting those damaged photoreceptors, researchers injected a viral vector carrying the gene for ChrimsonR, a light-sensitive channel, into retinal ganglion cells, the output neurons that normally relay processed signals to the brain, making those cells directly light-responsive. Patients wore goggles that pre-process the visual scene (shown as edge detection) before sending it to the eye, since ganglion cells expect compressed signals rather than raw light. The study's main goal was safety, with a secondary aim of measuring whether light sensitivity improved in the treated eye compared to the untreated one.

  1. 01

    The trial was open-label, meaning participants knew they were receiving the gene therapy rather than a placebo, which the hosts note limits how strongly results can be interpreted.

  2. 02

    The goggles' edge-detection processing is described as a crude approximation of the retina's actual computations, which include specialized cell types for motion, on-off responses, and other features beyond edge detection.

  3. 03

    The hosts contrast this approach with an earlier-covered device called Prima, which uses an implanted chip responding to infrared light rather than gene-based optogenetics.

  4. 04

    Channel rhodopsin, the light-sensitive protein used in the therapy, was originally discovered in algae by Peter Hegemann, one of the Nobel laureates discussed in the episode.

Transcript

2,978 words · auto-generated from the episode video

10:24one is going to be a direct medical application and the second one is going to be more kind of fundamental sciency and really the reason why these three authors won. But the first one, you know, um I think it's still pretty cool that you can directly apply it to a medical therapy. So this has to do with a vision treatment. This vision treatment asks the eyes output cells to take over the job of broken light detectors. effectively like there's a you know there's a disease called advanced retinitis pigmentotosa where the retinal pigments that capture light from the surroundings they get deteriorated. This is an attempt to restore vision among those patients. Okay, it's a paper from the University

11:05of Pittsburgh and actually really a whole consortium out in the New England Journal of Medicine. Again, I had to get my bootleg version from my friend who's a doctor. Um but this was out in October 2026. So like this month, right? Very very recent. >> Um and it examines a optogenetic therapy in 10 people with blindness due to advanced retinitis pigmentotosa. So to understand this approach, let's first actually just review what the retina is as a neural circuit. So this is what the retina looks like. The retina is one of my favorite neural circuits, not only because of its simplicity, but also because um in undergrad at Princeton, my

11:46senior thesis was on physically modeling the retina as like an electrical circuit in some sense. It's a very simple circuit. Okay. In the back of the retina are the rods and cones. This is the stuff that's in the back like towards the back of your head. So the eye the the light is coming from the center from from in front of you. the sensors are actually in the back, which is weird. You would think that the sensors, like in a traditional CCD, it goes CCD and then the wires come out from behind it, right? Um, here it's the opposite. Like the the the CCD is kind of in the back and the wires come out forward and then have to go through a hole to get to the brain. The reason for that is because

12:27our detectors, the rods and cones, they deteriorate very quickly because they respond to so much light, right? the proteins are always like, you know, changing shape. That's going to lead to degradation. And so you need to be close to blood vessels to like trash all of the proteins and get new proteins or get enough energy to get new proteins. You know what I mean? >> Yeah, 100%. >> So you need to be close to like the trash shoot. >> The the structural shape is related to how it's actually used functionally and proximity to basically the recycle mechanism matters. >> Yeah, exactly. Like if if the rods were actually in the in the same design of cameras, then you would just get a buildup of like your rodopsins that have been used up in the front and you'd very

13:09quickly not be able to see, right? So, okay, that's why the rods and cones are in the back. And then you have a bunch of cells in the middle that are kind of like processing cells. There's bipolar cells, amocrine cells, and at the very end you have these things called retinal ganglen cells. These are single cells that have their cell body in our eye and they've got an axon that goes through the hole that I was talking about from the retina because they're in the front. So, you know, you got to topologically go through the hole. There's a hole that goes through the eye to the to your brain, right? And it crisscrosses right in the middle and that's where the center for your um uh the circadian rhythm is, all that other kind of stuff. But in any case, the the actual sensors

13:52that go all the way to your brain are in the very front. Okay? Those are the retinal ganglen cells and the cell bodies are in the very front. There's a little bit of processing that's happening. I want to I want to be clear here. This is not just simply whatever I see in this pixel is what I'm sending. There's a little bit of data compression that happens within our eye because you can't send all of the raw data, right? It's a thin wire. Um you better not fry it and you better not heat it up. So you have to sort of compress the data in order to send it to the brain in in a efficient manner. And the retina is doing that. >> Our eyes are pied piper for data compression. >> Yeah, dude. They're creating zip files in some sense. The the the retina itself

14:32is creating a zip file to send to the brain. >> It's pretty remarkable. And I love this this >> network for that reason. It's very simple. There's only like five or six different types of neurons involved and yet you can have this rich machinery for a lot of very cool computation. And everyone can understand the benefits of having eyes because it's a part of our, you know, core five senses and our lived experience um in terms of how we interact with the world around us. And so to have something that's so visceral to everyone's lived experience >> and the the mechanics, the functional mechanics being >> complex but in a way that is simple enough compared to other biological systems uh is is actually

15:14quite interesting. >> Exactly. Yeah. And the the problem that these patients have with um this retinitis pigmentotosa is that the photo receptors the original photo receptors get damaged. Okay. So how do you do this? Well the strategy is to give retinal ganglen cells a microbial light sensitive channel. >> You basically do optogenetics on the cells in the very front. The ganglen cells that are sending messages to the brain. Okay. So you're like, okay, I'm not going to care about the photo receptors and things like that. I'm not even going to care about the middle circuitry. I'm just going to go to the output. I'm going to express my optogenetics in that. And then, you

15:57know, when the when when the things light up, the light is going to go in, but it's not going to hit the rods and cones because the rods and cones are kind of degraded, maybe even dead. The light is going to go hit the channel rodopsin that's going to cause these ganglen cells to fire, which is then going to send the signal to the brain. >> Ah, so I see what you're saying. So the the the middle and back layers, right, are not there's a direct connection between what's happening in the ganglen cells in the front that sends the actual signal to the brain that does the interpretation. And so even if your middle and and back uh machinery is degraded, which is what's happening in this particular disease, >> we're just overriding that by saying we're we're going to take the direct route. We're going to take the express lane uh and activate what we need to in the actual sensors that are taking in

16:37the light and doing this zip file type uh compression to the source. Exactly. That's exactly the point. And so how do you actually administer this thing? Well, in the 2026 trial that we're talking about here, each participant received an injection into the worst seeing eye. So they had like two eyes. One of them had this retinitis, the other one didn't. So in the worst eye, you inject this kind of capsid, which is like a viral capsid with this particular gene. There's a genome for um crimson R, which is the um the like the channel redopsin type thing. And then you've got a promoter that like kind of tags the retinal ganglen cells. And once those cells see this piece of DNA, they can

17:19express the protein on their membranes and then they are now light sensitive. Before they weren't. >> This is the delivery device. >> Yeah. This is the delivery device. Exactly. And so one key point to make here is that we're going not to the photo receptors but to the ganglen cells. Right. And I told you that there's some compression that's happening, right? So you can't just deliver light like it's normal eyes, okay? You have to use specific types. You you have to do a little bit of the data crunching >> before you send it in, >> right? Because the ganglen cells are expecting a data crunched >> set of numbers, right? A set of stimuli.

18:01They're not expecting the the raw light. They're expecting all of the they're expecting zip files. >> Right. Right. Right. Right. >> And so and so you have to you have to give them kind of zip files. >> We have to pre-process. >> Yeah. We have to pre-process. And that's why we need goggles. Okay. So there are goggles that the patients wear >> that are that's not going to actually give you the the raw image. So on the lower right you're seeing what the projected image is from the goggles. Okay. In this case what they're doing is like kind of an edge detection. Okay. the the left is just a hand, black and white, but the edge detection is detecting movement and like changing pixels and it's only highlighting those pixels. Now, one thing that the authors are careful about is that this is not

18:42the only type of pre-processing that the retina does, right? The circuit is extremely complex. There are onoff cells, there's movement cells, there's all sorts of like different types of um processing that happens and this is just one of them. Um, and furthermore, in the ganglen cells population, there's specialized ganglen cells that do one thing or the other, right? It's not like all of them respond to ed edge detection. In a specific slice, there's going to be some edge detection cells, but there's also going to be some onoff cells and so on and so forth, right? So, this is a crude way of kind of approximating what the retina does, but it's not even close really to what the retina does. The point of this study was

19:22to establish safety, >> okay? so that they can go on and do other studies, right? Because you're doing a lot here. You're administering some genetics into a patient's eye. You're trying to see whether it responds to light. You're trying to make sure that it doesn't give them like epileptic seizures in the worst case scenario. You know, there's all sorts of things that can go wrong in biology. So, the first and foremost um primary purpose of this 2026 study was safety. The secondary one was to measure okay like does the full field stimulus like change the threshold of the eye? Can the eye like increase threshold sensitivity? This is why they actually did patients with one eye with retinitis and the other one normal because then now you can compare like

20:03how how much better is the left eye compared to your control which is the normal eye, you know. Yeah, it's it's a well it's a well done study. Um there were 10 patients. >> That's really clever actually. >> Yeah. Yeah. And it was it was done pretty well. The 10% study open label. So that means that I mean obviously the people know that they're getting the viral vector, right? There's also if you wanted to do a real study, you'd have to do a fully blind one where people are getting saline, right? And then I don't know what if the the mind body connection I see better. Yeah. Like it's things are crazy, dude. The mind is crazy and like that's totally an outcome that could happen. So um this is the first in in in multiple. >> Mhm. >> But it's it's it's a pretty cool

20:45scientific payoff, right? It's a new route into surviving neural circuitry. One of the things about this story that you know because when we talked about the prize originally, it was just looking at the sort of the past and how it built up to now this being again an October study >> that came out literally like >> like this month, >> right? You know, just not that long ago. Um and how we've taken what the original an original question which was totally unrelated to giving partial uh visual restoration and we can say like in the clinical context we can say okay well there's only partial vision repair and we talked about this ed edge detection idea like okay well that's you know

21:25you're not giving people their sight back and it made me think about this idea of you know >> for people like us that currently maybe not into the future currently have our full ability ability to see in our site. It can be hard to understand the impact of even this first incremental step in someone who was a part of the study and is experiencing it directly. We can look at the measurements. We can talk about the calculations. But there is a real lived experience that is getting fundamentally changed. >> Yeah. >> Even if not perfectly back to their original state by these early steps. Um, which I I find, you know, it's important

22:06to always remember that when we talk about the these things ending up in patients, uh, these are people who didn't even think there was an opportunity for them to get their sight back. >> Um, and I think like some of the things now it's like in Yes. Maybe they can't make out a person's face. Maybe they can't, you know, uh, read perfectly, but they can start doing things that were removed from their lived experience. That I think is is really really um, unbelievable. Again, we started this by looking at like how uh it was work in algae. >> Yes. Channel rudopsson was first found in algae by Peter Hegman >> and then it became oh this is a way for us to be able to uh create and drive

22:47action within neurons. >> Mhm. >> And then now this is being applied by clinical researchers for restoration of eyesight. Mhm. >> And when we you were looking at this idea in algae, you were not thinking that, oh, this is going to lead to restoration in eyesight. Maybe once you got to the transitional step of understanding that you can control neurons, obviously the that's Pandora's box and but um >> some of these things can take >> one of the things that I think is incredible about this discovery especially is >> sometimes when you're these researchers, you don't get to live to see the impact on people's real lived experience of the work that you do. >> Yeah. uh a lot of these researchers have been a are still around and are able to

23:28actually see it come to fruition in everyday lived experience which sometimes basic research and lived experience there's a large delta between these two things I think that's just such a cool part of this story um >> yeah and it's only going to get better right yeah like this is the first >> few things I mean this isn't even the first um paper that we covered that was trying to restore >> eyesight we had the prima paper earlier But and that that one was actually much more involved because that involves like a a literal chip going in the back of your eye that is responding to infrared. >> Great episode. That's a crazy episode, too. Yeah. So, I mean, you know, >> we live in a we live in a day and age

24:08where like medicine is the frontier, right? It's >> it's it's pretty cool the things that we can do. >> And there are more than one ways to solve the problem and this just happens to be one. But I think the thing you brought up that's important is uh safety was the key aspect of the study, but they now have uh a foundational platform around the those um was it uh uh ganglen gang >> on the front >> knowing that there's multiple functional pieces to it. Okay, now we can start to take this uh success >> with doing one to now making it a a multipplex kind of solution and seeing if we can now not just do egg dete edge

24:49detection. Can we get movement? Can we get these other things? It's going to be fascinating to see. And a lot of people in the comments were like, I'm studying optogenetics. We do it, but we're doing this and we're doing that. So, it's an area clearly of of robust interest in research. >> Yeah. I mean, as I said, every single university has an optogenetics lab, right? Um, the main justification really for the optogenetics prize is is not its application in these niche sort of medical disciplines, which of course it's amazing that it's happening, but the real bread and butter for why optogenetics has transformed neuroscience is because of what I talked about earlier, which is this ability to establish causation between neural

25:30circuits and behavior. neural circuits and aggregate bodily functions in some sense, right? Um it's become the hammer for neuroscience fundamental research

From the episode
  1. EP 64

    Nobel Prize 2026: The Science Gets Stranger

    How Nobel-winning science leads to vision research, memory-controlled metabolism, cosmic neutrinos and sensors for mirror molecules.

    Nobel Prize 2026: The Science Gets Stranger

ChemistryNeuroscienceNobel PrizeParticle PhysicsAstrobiology