 And I think we should start right now. So I'm Jun Yang. I'm associate professor at Moran Eye Center. So today I'm going to introduce our speaker at this Grand Rounds. So today we are honored to have Dr. John Flannery here on campus to give us two talks. One is this Grand Rounds and the other is our department distinguished speaker research seminar at noon. So if you have time, you're welcome to join us at noon. So Dr. John Flannery is currently the professor of leading science, molecular and cellular biology and neuroscience at Helen Will's Neuroscience Institute at UC Berkeley. He obtained his bachelor's and PhD degrees from UC Santa Barbara, where he conducted a study with Dr. Steven Fisher doing research on circadian disc shading from photoreceptors in Xenopus retina. And then he did his post-op training with Dr. Dean Buck and Dr. Deborah Farber at the Julie Sting Eye Institute at UCLA. And during this time, he developed the first human RPE cell culture system, which is still in use in several laboratories today. And also he conducted one of the first transgenic rescue experiments for early one months. In 1991, he studied his faculty position in the Department of Thermology at the University of Florida. And in 1994 or five, he moved his laboratory to UC Berkeley. So the long-term research interests of Dr. Flannery are the genetic and biochemical aspects of inherited retina deteriorations and develop treatments, gene therapies, for this group of diseases. His lab studied the normal retinal function and the disease species by developing and characterizing small and large animal models and also design and develop various viral vectors and promoters to target different retinal cells. So they recently developed optogenetic gene therapy by delivering optically-switched receptors and ion channels into retinal ganglion cells and bipolar cells. And Dr. Flannery is right now a vice chair of Scientific Advisory Board of Foundation Finding Blindness. And he was also once the associate and acting director of Highland Mills Neuroscience Institute at UC Berkeley. So Dr. Flannery had published a model of 150 papers and received numerous awards, including the board of director's awards from Foundation Finding Blindness. Now let's welcome Dr. Flannery, so the title. Thanks, June. Thanks for coming. This is clearly not the title. Thanks for inviting me. I changed the title from your program. What I'd like to do this morning is in November, I was invited to Roche in Basel for a set of talks where they wanted to have about 12 people come, different clinicians and basic scientists, and tell them about where they should be or where they should try to be in the next five years. And this is the Roche that owns Lucentus. So I kind of felt like they were doing OK. But they said, you know, we were sure that injecting Lucentus every six weeks in patients is not going to last forever. And so we want to start thinking about what to do in the future. So of course, since they're Swiss, they had to change the title of my talk, something that they like and put their logo on it. But then they said, as soon as we got there, they arranged the meeting so that everybody sat in a circle. You had to wear a headset to talk. It was very much like a TED talk. They were trying to be really hip, even though they're Swiss. And then they said this thing that's a very loaded term in San Francisco. They said, we really want you here the next day and have to be disruptive. And when we think of disruptive in the Bay Area, we think of three first-year residents like in a garage making an app. And you know, this is where we are, right? We have two 40-story buildings. They have thousands of people. They're like the least disruptive people on the planet. Anyway, I thought for the next half hour, I'll give you the talk that I gave as sort of where when ophthalmology may be going as far as the kinds of projects that we work on in my lab over the next couple of years. So one of the things that's become quite clear over the last couple of decades, particularly in my role working with the Foundation Fighting Blindness, is that initially Ted Drysha and Peter Humphries found in the 80s the first gene that was causative for retinitis vivintosa, which is a P23H mutation in rhodopsin, which is kind of geared in my heart because it's an Irish immigrant that was the founder that came to the United States. This is not found in the UK. The first patient with this is American. Anyway, everybody thought, oh, we found the first RP gene were finished. People that do medical retina have been a little bit too good at this, so now they're up to 265 different causative genes for retinitis vivintosa. And Steve Dager in Texas has to maintain a webpage which basically changes every day as people find new causative genes for inherited eye diseases. The good thing about this, if you think about it, is that this looks like it's actually plateauing and people that do medical genetics say that this is probably all the genes for most of the large families. There's still gonna always be small families that aren't on this list and there are always gonna be new random mutations, but the numbers 275 is probably somewhere in the ballpark. It's not gonna be twice that. Now that we know a large number of the genes of patients that come into clinic with progressive blinding diseases, you can start looking at what they encode, and what you can see is that they encode almost every biochemical process in a basic biochemistry textbook. You can find mutations in mitochondria, in ciliopathies for making structural proteins of ciliaphoid metabolism, almost anything that you can think of, there's a set of patients that have a mutation in that. And so all these together sort of lead to the same clinical phenotype is that the patients lose photoreceptors. These mutations kill photoreceptors. And so what you see in a combination between looking at the genetics and looking at the biochemistry and then looking at really high resolution imaging is that most patients have mutations that are rod specific. Their rod photoreceptors die first. There's a few cases where the patients have mutations in cone specific genes. And those are like achromatopsia, blue cone monochromacy, red, green color blindness, but almost all the rest of those 265 are rod specific or rod killing mutations. So what you see in the clinic is that there's a few cases. One is congenital stationary night blindness is very rare as a bipolar cell. There's a nystagmus that's in amicron cells. glaucoma is probably something to do with ganglion cells but the rest of them are up here in the rod photoreceptor layer. And so what you see in a very, very late stage disease patients with inherited retinal regeneration is that they end up with basically no rods and no cones but the imaging suggests that they still have the second and third order neurons, the bipolar cells and the ganglion cells for decades. So what I'm gonna do for this talk is explain something about the first part and then at the end, we'll talk about therapies for patients that no longer have any photoreceptors. So when I was a postdoc at Jules Stein, we got a few donor eyes that were quite informative and this is a 17 year old patient that had a retinitis pigmentosa mutation. You can see this is the typical appearance that you see in inherited retinal regeneration patients that they have healthy RPE and coroid and in the central retina, here's some macular and the fovea, they have quite intact if not normal foveal cones and macular rods. And then you can see as you move out of the macular, you see progressively more and more degeneration of the photoreceptors. Here's photoreceptors with short outer segments but there's good RPE and good coroid and as you move up further, you can see that there's highly disorganized photoreceptors here where there's just one short outer segment. So most of the patients for most of those mutations end up with some progression like this. Now there's a few examples where the disease mutation is in the retinal pigment epithelium. One of those is corioremia and this is a case of a woman that was a carrier for corioremia so she had patches of area where the corioremia gene was expressed in other areas where it wasn't. You can see here's an area where the RPE is normal and you can see normal photoreceptors. Here's an area where the RPE is abnormal and we have very truncated photoreceptors. Here's sort of an edge of one of these punched out areas where there's photoreceptors until there's no more RPE and then there's areas with no RPE and no photoreceptors. So this is a very rare situation but this is a defect where the retinal pigment epithelium has the disease mutation and it kills the photoreceptors as a result of killing the retinal pigment epithelium. So this is actually quite rare from what we know of what the genetics of the diseases are. Okay so spark therapeutics got the first gene therapy approved last year for any disease for labor's congenital lamerosis type two. When you talk to patients they are scratching their head especially if they're in a big meeting with lots of patients is why in the world would you pick that because my family doesn't have it or it's so rare or whatever. It actually is quite rare and it's quite a bit not representative of what I just told you and that LCA2 is a defect in the retinal pigment epithelium but unlike caroteremia it doesn't kill the retinal pigment epithelium. It's a defect in the vitamin A cycle and in those patients one of the reasons that it's a good choice for gene therapy is that they have intact pigment epithelium and intact photoreceptors for decades. That gives you a very long window to treat those patients because the cells aren't dying and they have a huge sensitivity loss because the defect is in the enzyme that makes 11-cis retinal and that is the photopigment chromophore for all the cones and the rods. So the patients have such bad vision they have nystagmus and they have very early loss of sensitivity because they can't make any chromophore because their defect is in the enzyme that makes 11-cis retinal. But also if you put this back and these patients are recessive, null for it so therapy is to give them a normal copy of the enzyme to make retinal they get a 5,000 fold increase in sensitivity in a couple of weeks because the cells that make the 11-cis the RPE are intact and the photoreceptors are largely intact they just don't have any photopigment. So what happens here, this is a visual cycle and the defect in those patients is in this enzyme here that makes 11-cis retinal comes out of retinal pigment epithelium is transferred to the photoreceptors and then after you hit it with light it turns back and goes back to the RPE and is recycled again. So these patients have a defect in this. So even though it's a very good candidate and the patients get much better quickly it's actually not representative of therapy that's gonna work for most of the patients because as I said at the beginning most of the patients having defects that are killing their rod photoreceptors later they're losing their cone photoreceptors they're not having defects in very few cases in the RPE and then one of the other conditions, choroteremia the defect is actually killing the RPE. So this LCA2 is actually pretty anomalous. The actual reason that was the first gene therapy for ophthalmology or anything else is Soros serendipitous. In my previous job when I was in Gainesville Bill Houseworth was in the lab next door to me and he got a call from Gus Aguieri who's a veterinarian at UPenn and he had a dog that was bred as a show dog that spontaneously got LCA2 and so that was really the instigation of this project. It wasn't designed by a company it wasn't designed to treat the largest number of patients or have the best phenotype. It was pretty serendipitous that the dog appeared with that mutation and that the characteristic of disease is actually quite amenable to gene therapy. So here's a slide from Sam Jacobsen that suggests if you're gonna do gene therapy for other diseases maybe the ones that are more typical of what you see in the clinic it depends where in the state of the degeneration the patient is as to what therapies are gonna be appropriate for them. So in the case where the retinas normal looking where they have a full complement of rods and cones you probably have more options and one of the options would be like in the case of LCA2 with Spark you identify the gene defect in the patient in that case it's the enzyme RPE65 and you put back a normal copy you correct that one gene defect as long as all the cells are still there and you can put back that one gene that's probably the best therapy no one's gonna beat that with anything else. However as the patients are in later and later stages they're losing photoreceptors some stages the retinas getting edematous and getting thicker it's ultimately getting thinner at the very end stage they don't have any photoreceptors at all they have no rods and no cones so no matter how good your gene therapy is and how well you know exactly what's wrong with these patients you can't do gene replacement this is the cells you're gonna do the gene replacement for in this case they're not there anymore. So I'll talk about a few examples of therapies that we're working on to try to treat patients that are in the early stages and at the end if I have time we'll talk about a new field of optogenetics with the idea is to add a new gene to these second order neurons that look like they survive clinically but they're not light sensitive. So as I said the gene therapy for almost all the 260 something mutations is gonna be targeted primarily to rods because that's where the site of the mutation is in almost all cases. So what you're gonna wanna do is you're gonna wanna put the therapy between the photoreceptor layer and the red lumbar epithelium. And so what you can see here no matter what stage you're at whether your idea is that the patient has large numbers of rods and cones and you wanna slow their therapy by either giving them stem cells or if you know what the gene is and that you have a tool to deliver it and you give them back a copy of that gene or if you wanna give them new photoreceptors and new RPE cells in many cases what you need to do is you need to do surgically a sub-retinal injection because you have to put either the virus or the cell in the sub-retinal space. And in the spark trial all the patients are treated by sub-retinal injection and there's about 10 other companies that are doing clinical trials in various stages and different clinics and all those with one or two exceptions are sub-retinal surgeries. And so what you would like for some examples is to switch that from sub-retinal to an intervitual approach. One of the reasons for that is that it would be surgically much easier. It's more akin to what you do with a Lucentus injection. Also it gives you a much bigger area of the retina that you can potentially treat and the surgery's gonna be much less complicated. Okay, so one of the challenges is the viruses that people are using for clinical trials particularly the one that Spark uses were not isolated from ocular tissue. They were circulated from patients from serum or from different systemic tissues like the liver, the kidney, or the lung. They're not necessarily optimized to transfer the gene that's in that virus to retina. What's nice about it is that in the case of Spark if you put that virus in a sub-retinal space it's actually quite good at transferring the gene for RP-65 to the retinal pigment epithelium. But it wasn't engineered, it's really just serendipitous that it does that. But an example for almost all of these where the patients are recessive patients where they're missing an enzyme or they're missing a protein and you know exactly what it is and it's only one gene, the therapy will be the most elegant one would be to replace that and that's what Spark does. And in those cases almost always the target's gonna be rods and cones except for in the case of LCA2 and corioremia where it's RPE. And in glaucoma it may be retinal ganglion cells. The jury's still out for what the target and the mechanisms are for glaucoma. It turns out for a secreted factor there's a couple groups that are trying to do the genetic version of Lucentus and Ilea. So S-split is a soluble VEGF receptor. It looks like it doesn't really matter because like you're doing with an injection of Lucentus into the vitreous all that matters is the amount of the dose if you will of the protein. And if it's secreted it doesn't really matter what cell it's secreted it from. It matters how much and where it goes in the retina. So in this case it probably doesn't really matter if you get it in the RPE or the photoreceptors or some other cell. And you probably wouldn't want to do a sub-retinal injection for this because it seems like it's an unnecessary surgery to make a detachment. So fortunately the gene therapy trials for LCA2 and the other conditions are actually going quite well. Now there's 244 is probably more than that by this week that have been injected with different virus vectors in the different trials. So far there's been not a single what FDA calls an adverse event where it was uncontrollable uveitis or a tumor or no one's been enucleated so far. So it looks like it's quite safe as far as using viruses in the sub-retinal space. And the FDA is getting more comfortable with this idea of doing a sub-retinal injection of a virus to deliver a gene for an eye disease. The interesting thing is the FDA, their outcome measures that they accept for these trials is remarkably low tech. They don't allow OCT, they don't allow electrophysiology, they don't allow visual fields, they allow acuity and Spark has built an obstacle course which you've probably seen where the kids navigate through a dimly lit room. So the FDA does not like for these trials most of the things that you use in the high tech clinic as a readout, they want the lowest tech readout as possible maybe because they're more reproducible in different centers. So they don't allow any of these things for the LCA2 trial, just sort of surprising. So there's quite a few that are ongoing that are not approved. Spark has actually approved for commercialization for LCA2, you can write a prescription for it now. The biggest concern with that when you talk to patients is the price where it's $800,000 per patient. Coroid oremia is in the phase one, two in several different centers and there's several other genes that are in several different stages of phase one, two. And one of the things that's surprising is that some of these incredibly rare conditions have multiple clinical trials run by multiple companies for the same very rare disease. And many cases there's absolutely none of the 250 genes that are known that anyone's looking at. And so the reason for this is sort of the same as what I said at the beginning is that these are not picked on the basis of their prevalence. They're sort of picked at the ones that they have a gene that fits in the virus and that the gene is known and the size of the gene and the way it works biochemically is understood. To some extent these companies are running on the basis, a trial and picking a disease on the basis of what their toolbox fits, not what the patient need is. So it goes on, there's more and more. What you see for RPE65, there's more than three companies that are working on the same disease. The projections are there's maybe 2,000 or less patients with RPE65 defects in North America. Okay, so one of the things we did in the lab for the last couple of years is try to see if we could re-engineer the virus that people are using. Everyone's using adeno-associated virus. What's good about adeno-associated virus is incredibly safe. 80% of the population are already serum positive for it. You've been infected for somewhere in hospital by it. It doesn't cause any pathology in the normal case and it looks like in many cases it's good and in the case of LCA2, it's good at transfecting retinal pigment epithelium and just the serotype that naturally occurs. However, one of the things it doesn't do is it doesn't penetrate through the retina from the vitreous and that's the reason for the surgery requiring a sub-retinal injection. So what you see is that if you take unmodified AEDV, serotype two or the other naturally occurring serotypes one through nine, if you take the normal eye, whether it's a mouse or a macaque, and you inject those particles into the vitreous, they only go to the first layer of cells. You see them very efficiently transfect retinal ganglion cells but you see no viruses in the middle of the retina and absolutely none in the photoreceptor and RPE. If you put the same virus in the sub-retinal space with a blab, what you see is they're very good at infecting photoreceptors and RPE but by the same token, they don't travel backwards towards the ganglion cell layer at all. And so here's some examples that we did early on just to show the difference. If you take one of these viruses, this is AB2 and you put a green fluorescent protein and that's so you can see where it goes. If you put it in the sub-retinal space, you can see it basically makes expression of the gene of interest. This could be the disease treatment gene but here's the fluorophore. It makes the treatment the size of the blab and the locations of the blab. So when a surgeon's making the blab, the blab is sort of random shape and it's in some cases a random location because a surgeon may put the needle here and then blab may migrate towards the periphery. If you do the same virus with the same volume with the exact same gene, you do it in the vitreous, you see you get expression all the way out to the aura but in this case these are retinal ganglion cells. In this case it's photoreceptors and RPE. So that's what's diagrammed here and in most of the patients what's happening is the treatment in the clinical trials is a sub-retinal injection of under 200 microliters into the sub-retinal space between the photoreceptors and the RPE and what you're doing is you're putting the particles in contact with the cells you want them to treat. And the reason for that is that these viruses look for cell surface receptors and the receptors they look for, the naturally occurring ones are really common because the wild-type version of those virus, their goal in life is to get as many cells as your body as possible to copy themselves and give you influenza or whatever they're encoding. And so they're looking for very common receptors like heparin sulfate, gloria glycans and the FGF receptor that in the retina are very, very common on the interlumining membrane. And so what you see is that if you put them here they basically, it isn't that they're not small enough to percolate through the retina. Initially people thought the size of the particle was so big even though they're only 20 nanometers they just couldn't get through the spaces between the cells. It's that they find so many things to stick to in the ILM and the ganglion cell layer that they basically get used up if you inject them in the vitreous and the same token if you put them in sub-retinal space is the edge of the initial bleb that the surgeon makes is the virus doesn't go any further than that bleb and it doesn't go from the sub-retinal space back into the retina at all. So we did a couple years ago, we tried to make libraries. We made millions of different AAV viruses by making every possible substitution in the proteins on the surface and we wanted to screen them. And these are ones that didn't occur in nature. We wanted to look for properties that were attuned and specific for our needs and ophthalmology. And that in this case is to penetrate through the retina from the vitreous and get to the RPE and the photoreceptor layer. So what we did this is we took the gene that encodes the outside surface of the virus and that virus has 52 copies of three different proteins. We made every possible change that you could make in it. So we made mutations, we made what we call shuffling. We took different viruses and we cut them up and shuffled them together. We also put peptides on the outside and we took these libraries and on paper this is 100 million different AAV virus variants that don't occur in nature. And then we had to design a screen. And so in our case for the eye the screen is to inject the libraries of all these millions of variants into the vitreous and then wait a few days and then collect the retina and look for which ones that won in this competition for getting to the photoreceptors in the RPE. And what you can see is that as you do this what we did is we injected into a monkey or a mouse. We wait a few days, we isolate the retina, we take out the winners if you will then we do it again. We go around and around and around and we narrow it down from the 100 million to a very small number. What you can see is that about round three they start to converge and you get some that are very specific for retinal ganglion cells and you start to get some that are better at penetrating through the retina and getting to the RPE and photoreceptors. So we have two goals. The viruses have to get from the vitreous to the sub-retinal space without a sub-retinal injection but they also have to still be able to infect the targets which are photoreceptors and RPE. So that's what's shown here. You make the library, you inject it into the vitreous, we've most recently done this in macaque. You isolate the retina, you take out the viruses that have made it to the outer retina, you select those few, you do it again, you keep doing it again, you pick a narrow and narrow subset and you go back around again. And so one of the best ones that we found from the vitreous will transduce the entire macular infovia from the vitreous side with no sub-retinal injection. What we see is you get further outside of the central retina, that's not as good than the interlimiting membrane in a macaque, is probably the major barrier that we need to conquer. But for this, we see that we're able to transfect the entire infovia and macula. So you can see here, here's one that was optimized for retinal ganglion cells. So here is the retinal ganglion cell output of the macula and infovia. And here's another one that shows all the foveal cones transfected from an intravigal injection. This is 50 microliters of one of the best variants that we selected. And in cross-section, you can see it's very efficient in infecting the retina and the central retina. You can see these areas here, which I showed you as the spotty areas in the fundus picture. What you see is where there's a retinal vessel that's against the inner retina. It looks like the retina's thinner and the viruses will get through there. And in the areas between the vessels, it still has a big barrier by the interlimiting membrane. So it looks like the IOM barrier where there's a blood vessel that touches is actually less of a barrier for the virus. We also found one that infects Mueller cells and no other cells. So for some applications, like secreting S-split, for example, for a new vascular AMD, this may be a good target because it's a radial glia that runs through the retina. It goes from aura to aura, and you can easily get to the end feed of Mueller cells from the vitreous. So another therapy that's coming on for patients that are not recessive, all the patients that are being treated in those trials I showed you on the previous slide, they have recessive RP in that they're missing some component. So patients that have dominant usually make a toxic protein. And so giving them a normal copy is not gonna be enough. So at a toss, for example, one of their first targets for CRISPR gene therapy is CEP290, another rare inherited retinal degeneration. But instead of putting back CEP290, they've generated this virus which has a Cas9 CRISPR which will take the mutations out of the Cas9 gene and swap them back to the wild type sequence. So they're recruiting patients now. The virus they're using is AAV5, so they're going to use subretinal injection approach for this. And one of the issues is gonna be that the area that's treated is gonna be the size of the bleb. And the surgeon is gonna be nervous when they're making a really big bleb and maybe even more nervous if you're detaching the fovea and macula. For other diseases, CEP290 approach with one virus doesn't work. And so there's a group that's trying to fix mutations that were dops. And for that, they need to make two viruses because CRISPR is actually too big to fit in a single AAV virus. Another project that we've been doing for a couple of years is in collaboration with Jose Sahel, who's at the Institute of Vision in Paris. And what Jose found is what I showed you at the beginning is that we now know that almost all of those mutations kill rods, but many of the patients ultimately go blind. And the gene defects in the cones are pretty much only color blindness. So what's happening is that they found is that genetically normal cones require a secreted factor that they get from rods. And that's called rod-derived cone viability factor. And what happens is that the rods secrete this factor and it controls the glucose uptake of cones. And when the rods finally get to about half or two thirds of them missing, you start to see patients that are losing cones even though the cones are genetically normal. The reason for that is that they're losing this factor that's secreted. So the therapy that we've been working on is can you just put back the rod-derived cone viability factor, even though it's not gonna help you keep the patient's rods around, can you keep the cones around by supplying this secreted factor? So the diagram of how it works is that rods secrete this factor called RD-CVF. It binds to a receptor that's only on cones called Basagen 1. What that receptor does is it controls a glucose transporter. So the reason the cones die in these patients is that even though they have plenty of glucose in the sub-random space and you can even inject glucose, it doesn't work because they won't transport it in unless they have RD-CVF. So when the rods are gone, the source of the RD-CVF is gone. So the cones are actually starving in the presence of plenty of glucose just because they won't take it up. So the therapy is can you just add RD-CVF and secrete it and keep the cones around? And so you can see here, here's a set of experiments where we put a virus with the gene for RD-CVF and the upper panel is eyes in mice that have inherited rod disease. You can see that the red is the number of cones. They're losing dramatic numbers of cones. You can see if you add back RD-CVF, the number of cones in the other eye of the same mouse is dramatically much more. You can see here the difference between the uninjected eye and injected eye. So even though this is far from a perfect therapy and that doesn't help the patient's rod mutation, it looks like if you supply this factor, the patients may be able to keep their central cones for quite a while. We've recently done this in a transgenic pig model that Maureen McCall made of RP. And you can see here's a number of cones in a wild type pig. Here's in the RP pig, there's basically just little nubs of cones. And here's an eye that's injected with the AAV for RD-CVF in the vitreous with the same mutation. So it's showing that most of the cones in a pig, surprisingly, pigs have more cones than people. I was surprised to find out. It's actually quite good at an intervitual injection will keep the cones around on this transgenic pig for months. Okay, so the last part of the talk, I'll talk about patients that no longer have the option of what we just talked about, where they no longer have any rods and cones. One of the therapies is to do what we call optogenetics, is that they have functional surviving inter-retinal ganglion cells and bipolar cells. And this is the basis for the prosthetic electronic chip that Second Sight makes, is that they put a set of electrodes hooked up to a camera on the patient's inner retina, and that that system, even though it's quite conky, actually does provide some motility vision for patients. So that shows that these cells are functional, they're just not light sensitive, and they're connected to the visual cortex for many years. And so what we thought is, could you use a virus gene therapy to add a new gene, not replacing a gene, a new gene that adds light sensitivity to the inner retina of these patients? And as these patients hear, they no longer have these options. So the idea is to do this instead of do this, where the patient has to wear a camera plus a grid electrode, plus a wire that goes out of the eyeball into the temple, and then there's a box with a battery and goggles. It's quite complicated. But it's not maybe in the inner retina, could you just add a light sensitive protein with gene therapy to the surviving inner retinal cells? And so what we found is that in blind mice, already one mice that have no photoreceptors at all, they have no rods and cones, the limitations of other approaches that people have tried is that the sensitivity for optogenesis has been so low that the patients that are tried in the retrospective trial, for example, they can only see at the brightest, like Florida, at the beach at noon intensities, 10 to the 15 photons per square centimeter. At any intensity less than that, they can't see anything. So what we wanted to do is to design a therapy that would work in normal lighting conditions, sort of like we're in now, and be fast enough so that patients could have motility. And so we're using the same viruses that I showed you, it's an intravitual approach. It's really amenable to an intravitual approach because we're trying to infect ganglion cells, not photoreceptors. So this is a relatively easy target for us. We're using some of the viruses I showed you earlier that we did from directed evolution, you can see here it'll get all the ganglion cells that are the output of the macula and the fovea with one transfection. And so we've been using, which many people told us wouldn't work, we've been using the cone opsin, the normal photopigment for middle wavelength cones, and transferring that to retinal ganglion cells. What we find is that it's sensitive enough to be functional in these animal models of the lighting that we're in right now. So what you can see as you express it, it's actually quite fast. This is a 472 nanometer light flash. You can see a response to the flash in ganglion cells quite quickly. It's fast enough to respond to 50 hertz, which we think is fast enough for motility vision. This projector is probably flickering at 60 hertz. So then we wanted to test how much functional vision can you get with this system. So this is an animal has no photoreceptors. We made a cage with a partition. We train the animal by giving it a mild shock to its feet. And it learns that if it can recognize the difference between a pattern, which is on a regular iPad on either side of the cage, it can move to the side of the cage to avoid the foot shock. And so we have an iPad with parallel lines that are vertical versus horizontal. You can see that after the optogenic transfer to ganglion cells of cone opsin, the animals can actually do this simple task as well as the animals that are normally sighted. And we found that you can do this task over three log units of adapting light. It's not nearly as good as photoreceptors where there are over 12 log units, but it does work in most indoor and outdoor lighting conditions. So it does have some adaptation built in. And more recently we built simpler tasks. This is just a box that has an animal with a bunch of toys and a ball, a triangle and a cube, for example. And what you can see is the untreated RD-1 mice that are completely blind. What they do is they basically skirt around the wall and they don't explore any of the toys because maybe they don't even know that they're there. What you see is a Rhodopsin-treated mouse, which we did earlier with putting rod opsin. They do more exploration than a blind mouse, but they don't do certainly a normal amount. Here's a wild-type mouse. What you see is they explore all the walls of the cage and all the toys and they make quite a complicated path. And the cone opsin mice actually have as much exploratory behavior as the animals that are wild-type, which is quite surprising. This is just a box on the bench in the lab in normal room lighting. So in summary, I like to say, I like to think that the virus gene therapy is getting better and that we're getting to the point where we may not require that sub-retinal injection approach very soon. So we've been giving out the viruses that we've isolated that work from the vitreous to companies as soon as they ask for them. And we think that in addition to the gene replacement, we'll start to see for dominant disease patients, CRISPR gene editing, like Editas's approach. And we also are hopeful that the optogenetic approach will start giving some idea of vision, some motility vision for patients that are no longer candidates because they no longer have photoreceptors. So I think I'll stop there. Thanks very much. Yes, hi. So this M-Obsing optogenetic model, have you tried how fast they darken out? How fast they darken out? Yeah, how they get the chromophore. Yeah, that's an interesting question. All the reviewers for our grant was wondering about whether there was chromophore. And I ended up having to go through the literature. It turns out nobody ever measured whether in any of the RD naturally occurring models if they still make 11-cis. So in order to resubmit the grant, we just had to do it. It turns out that the RPE in all these animal models that have no photoreceptors still make about 10% of the normal amount of 11-cis, which is more than enough for this system. So it generates. And when I talked to Ruston Gelder, who's an expert at patients that have these conditions, he said, well, of course, because they still have pupillary responses and many of them still have circadian rhythms and the melanopsin cycle requires 11-cis. So that's sort of hand-waving evidence that they make 11-cis, but we measured it. And it looks like they still make 11-cis for the life of the animal. So it could be interesting to measure how fast if you bleach and then see how fast they get the sensitivity back. Yeah, I didn't have time to talk about it here. I'll talk about it this afternoon, but when you look at these things on the multi-electrode array, you can start to look at those things because you have to add the, because the RPE's not there, you have to add the chromoform. Bleach is pretty fast. Yeah. So what is the incidence of retinal detachment with the subretinal injection? Well, if you talk to the surgeon, it's 100% and he makes the detachment. Right. Right? They all reattach. I mean, it's a question of whether or not, in diseases where you really wanna treat the fovea and macula, how much hydrogenic damage you do by detaching the fovea. Some people say none at all and other people say I would never do it. Paul Sieving's running a clinical trial for X-link retinoschesis and if you talk to him or Sam Jacobs on the idea of making a detachment with a kid that has holes in their retina already, makes them crazy. But when you think about it, there's so many holes in the X-link retinoschesis patients, there really is no such thing as a subretinal injection because it would come back to the vitreous anyway. So in the NEI trial for XLRS, it's intravitriol and AGCT is running a trial for X-link retinoschesis as intravitriol as well. If you talk to people that do PEEDS retina, they tell me that if you were to make blebs, you pretty much have to keep making more blebs because kids would keep coming back in as you fix them at one o'clock. They come back in a couple months later, you have to inject them somewhere else. So for many conditions, we like to think that the subretinal injection is not necessary. You just need a virus that will go through the retina. So it looks like we may have that. Yeah, Paul. So based on the economic success or lack thereof with Spark and their RB65, our company's more or less enthusiastic now about going through these huge processes and whether they're gonna agree whether they're actually gonna make money at it. That's a great question. We've been meeting with venture capital companies in the Bay Area about these things all the time. And at Vernon, which used to be Avalanche, their idea was to do wet AMD with actually one of the viruses that we made that they bought from the university. And their whole business model was that if you had to do subretinal injections, even though you may only have to do it once every couple of years for wet AMD, they didn't think they could do it with subretinal injections because you needed an OR and anesthetist and everything. So their entire business model was to do it like wasentus. The difference was on their calculation, $25,000 subretinal and $500 or something in trivaciril. So it kind of depends on the approach. So it'll be interesting to see many of these companies that are looking at conditions where there's only a couple of injured patients. I mean, there's not very many sub-290 patients. So, but Spark's price is not really based on what it cost them to get to the clinic. Estimates are probably cost them $200 million to get to the clinic. I think it's just based on what they think they can get away with. So, like Tim Stout told me, every patient that he did, he got reimbursed immediately. He said that Medicare reimburses stuff that they've never seen before all the time early and then they give them trouble for fecos and stuff, right? So I think for a while, right? It may be that they'll sign off on $800,000, but one of the other big issues is how long will it last. Spark has been saying that it's glass for the life of the patient. Many of the patients are eight years out. But Jacobson and Sedation say those patients are losing voter receptors at the same rate. So that suggests that it's not going to last forever. So, yeah, it's very controversial how the price, but the Lucentus price is not based on anything other than what Genentech thought they could get. Thank you so much, I'm sorry. Thank you so much for your talk. Just to make sure we have time for our second speaker, we'll move along and Dr. Plano will be together to talk today and you.