 is the director of the McPherson Eye Research Institute there, as well as the Sandra Lemke Trout Chair in eye research. Needless to say, he's very accomplished in both a clinical and surgical practice as well as his research practice. And he does really innovative work modeling some of the eye diseases we see in clinic, a lot of inherited retinal diseases and degenerations in stem cell models which are increasingly able to tell us more and more about these conditions that we've thus far been unable to really model effectively or understand on a molecular basis. So I am thrilled that he's able to talk to us today about some of his work. And I'd like to welcome Dr. David Gam. Thanks, Leah. So, I used to come here quite a bit. Those of you who've been here for a while, Ray Lund used to be here. And I had a couple of grants with Ray inside. It's a pleasure coming out to Moran on a regular basis for some time before he went to Portland. So it's good to be back. And it was nice having dinner with Leah and Matt yesterday, too. So what I'm going to do today is talk about a couple of... Well, first of all, I'm going to set the tone by showing you what we've done to this point to derive photoreceptor cells from induced pluripotent stem cells and ES cells, but mostly IPS cells to this point. And then talked to you about a couple of applications. And I know I have limited time, so I'm going to try and move along. But at the same time, if there are any questions along the way, just raise your hand. So it's a small enough crowd in a room where we can be interactive on this. Okay, some disclosures. I do have patents on the organoid technology that I'm going to be showing you here today through WARF in the University of Wisconsin. That ultimately led to the founding of Company Ops's Therapeutics, which is a subsidiary of Fujifilm to the Dynamics International, so bear that in mind. And then the main disclosure I want to give you is who actually does the wonderful work in my lab. And that is Beth Kompowski as a scientist and Joe Phillips also a scientist in my lab. Beth does a lot of the organoid work. Joe Phillips does a lot of the cell therapeutic work. And you'll hear their names quite a bit throughout the course of the talk. So the objectives of my lab, and I actually have my clinical practice as well, I pretty much a general pediatric ophthalmology clinical practice, but I do like to see patients who have genetic disorders and retinitis pigmentosa. And I marry that with the work that I do in the lab. But ultimately my goal is, and I exclusively use human pluripotent stem cell technology, but I want to use that technology first and foremost to investigate human retinal development. There really is no other way to study human development in a dynamic way except through pluripotent stem cell technology. You can get static images from fetal tissue when it's available. You can look at primates, but even that's difficult and expensive to do. So I set out in doing this because I had a strong interest in developmental biology. But then if we're successful in being able to use developmental biology to coerce pluripotent stem cells to actually become bonafide retinal cells, then we have the clay that we need to perhaps model diseases that we otherwise couldn't model from actual human patients that have clinical correlation or develop cell therapeutics. So that was what I hoped to be able to do, but when I started I thought, well maybe we might be able to get these things down the path a little bit. So we use induced pluripotent stem cells for the most part. We've developed over a hundred different lines from different patients and we generally use blood samples. And so patients will, we have a special kit that we use and we use T and B cells, so nucleated cells from blood samples. And we have a special kit that we send all over the world and it's about the size of a cholesterol, what you'd submit for a cholesterol check. And then we reprogram that using a combination of Thompson and Yamanaka factors, JB Thompson being at the University of Wisconsin, so we're kind of partial to those factors too. But at the end of all of that, you get an induced pluripotent stem cell which is capable of producing all the different cell types in your body, at least theoretically. It also harbors the entire genetic background of that particular patient, which is not a good thing if you intend to put it back into the patient eventually. You either have to correct the gene, which is a kind of, we can do that, but adds a very large level of complexity to the end of the whole process. Or we can take normal donor cells that are HLA matched or otherwise available to be able to produce the cell types necessary to target the disease. But if you're interested in studying the disease, using a human system, or developing therapeutics that are not cell based, so say gene therapies or pharmacologic therapies, then you can commandeer the fact that there is the genetic mutation still present. And if you're able to direct them towards the cell types of interest that are affected in the disease, you can then study and help test or design therapeutics. And we've done that with a number of different diseases at this point, and this afternoon I'll be talking specifically about best disease. But none of that works unless you can actually take those pluripotent stem cells which can make heart, lung, brain, whatever, and push them towards the cell type of interest and do it in a pure fashion because you can't do stringent assays if you've got some heart muscle mixed in there with a few photoreceptors. So you really have to be able to push it in a efficient manner towards the cell types of interest, pull those out, and then be able to do that reproducibly and have the proper controls. Otherwise, the rest of it downstream is just gonna be garbage. So it's not easy to do that. So if you start with a pluripotent stem cell, you have the ability to perform all the different major germ lineages. And then for retina, you gotta push it towards the actodermal lineage and then the neuroactoderm and then anterior neuroactoderm. You don't wanna make spinal cord. And there are steps in each one of these forks in the roads, there are cues. There's mechanical cues, there's chemical cues that you can use to kind of push them at various levels of efficiency. But ultimately you want the anterior neuroactoderm which includes the eye field, which has a lot of specific transcription factors that we can use to trace this as it goes down the line. And then from the anterior neuroactoderm you can develop forebrain or optic vesicles. Optic vesicles, a little bit of embryology here are imaginations of the anterior neural tube, one on each side that push towards the surface actoderm. This becomes the lens, the lens plot code. Ultimately this folds back in on itself and it vaginates, it becomes the bilayered optic cup where the proximal portion here becomes RPE and the distal portion here becomes neural retina. So even if you get to that point, you have a long way to go because at that point it has to make a decision was it gonna become RPE or is it gonna become neural retina? This is neural retinal progenitor cells. And we have the benefit in the eye that all of the neural retinal cell types come from a common progenitor. So you start with the same cell type. So all the different major cell classes within the neural retina. And indeed, even the RPE, if you go a step before, it comes from a single progenitor. So that's nice, we don't have a convening of a lot of different cell types from different cell lineages that have to come together which would be very difficult to do in the culture. But it's not just even as simple as making a neural retina progenitor cell and making all the different cell classes. They're made in an overlapping fashion over a broad period of time. So this is in mouse, so this is in the order of days so think about human where you're talking about months and months and you don't just get ganglion cells that are born first and then that gets done and then you move on to cones and you move on to the next one. You have these overlapping bell curves of cell types that are being made such that at any point in time you have a mixture of different cell classes at different levels of maturation. So a very complex system as opposed to say RPE which is a single cell type. And beyond that, you've got the complexity of the tissue structure. So it's not just that you're making them but hopefully you can make them with the proper spatial relationships to one another. And so here's just a quick diagram with all the different 10 major layers. So you're looking not only at cell types but hopefully synaptic layers. You're looking at inner and outer limiting membrane formed by the mucoclea. And here's histology kind of showing what this looks like in a human retina. Okay, so to skip over about eight years worth of work that's what we sought out to do in a very systematic kind of stepwise fashion to be able to take it from one step to the other. So first, can we make anterior and erected urn? Is there Paxix positive cells? There's this whole set of transcription factors that we needed to push it to. For that we had a lot of help from the neural stem cell biologists at the University of Wisconsin who were already moving this technology towards forebrain, spinal cord. So we had a lot of knowledge. We were able to stand on the shoulders of giants to get it to the point where we were at the threshold of the eye field or the optic eye field. And this required actually surprisingly little chemical manipulation. So we didn't have to do an awful lot to get it to go in the right direction but it required a lot of mechanical manipulation. So the key was to be able to pull these things up into a 3D aggregate. And so we did that in the very beginning when most folks were doing 2D cultures because we already knew that it was very successful in developing forebrain organoids. And as I showed you earlier, the neural retina is just an out pouch into the forebrain. So if it works for forebrain, I figured well it should work for optic vesicle too. And if I can get those optic vesicles then I've purified that step and then everything beyond that will be retina. So a couple of different steps. Ultimately we raise them up into three, we plate them back down and neural structures don't like to stay in the substrates on a plate. So these little areas, they make almost like little neural tubes with a little lumen in the middle. And then this area here is only lightly adherent to the underlying laminin or other substrate that we usually use multiple. And so they easily pop off either mechanically or we can use it with some trituration. And then lo and behold, we get these two sets of structures that are floating in 3D. We have these kind of more homogenous ones that are actually forebrain, I'm not gonna show you that, but the forebrain and the neural retina do grow together. And then we have these golden Cheerios, these vesicular neuroepithelial phase bright structures that are initially about the size of a head of a pen. And those can, because they look so distinct, we can separate those out, we can now do that robotically and get large populations of these structures. So what are these structures then? And then we can take them for upwards of even 400 days and allow them to mature like they would in utero. So this is an example of a bunch of these. We can get thousands of these from a single culture. And if we do a cross-section and we look to see what is it composed of, because I said before, there's a common retinal progenitor cell. So if this is truly an optic vesicle that can get rise to the neural retina, then initially it should be composed of a neural retinal progenitor cell. And we're very fortunate in the eye that they have a whole number of markers that specifically identify cell types, which other tissues do not have. So for in this case, visual systems, homeybox 2, and K-67 is a proliferative marker, basically show us that we have, the vast majority of the cells are starting out as a neural retinal progenitor cell. And if we let it go in this case, this is over 160 days, they change. We had a recent development paper that characterized the different stages that these went through. But ultimately they get to a stage where they're quite large, up to three millimeters. I mean, they don't get like an eyeball, but they get pretty big. Sometimes I'll have little bits of RPE on them, but they never fold back into an optic vesicle. I could talk about that more later, but they're actually flipped in terms of their orientation. And at stage 160 or beyond, we call stage three, the folks in the front can see this, but it has this very interesting, and this is all them from one culture. So you can see they have different sizes, some of them are circular, some of them are more oblong, some of them have modular aspects to them. And here's one that looks like a gingerbread man, so I always show this one. But if you look at the front, you'll see these little hair-like structures that surround the entire thing. And I'm gonna talk more about that later, but I'll tell you that those are outer segments. So they form an interphotoreceptor matrix, and they'll extend very long outer segments that are light responsive, and I'll show you that as well. And so we've gotten very good at this to the point where all of our lines, we can push a good majority of those organoids towards this late stage of development. So it gets you from where I showed you before, here's that, starting with that early organoid, if you take a cross-section, here are those progenitor cells, but the first cells that are born are renal ganglion cells, shown in purple. Then if you go a couple more weeks, then you start seeing the birth of early photoreceptor precursors that's shown by Cone-Rot homeobox gene here and recovering, and they undergo interkinetic movement. So they're going back and forth from the apical to basal side of this neuropathletal structure, dividing and setting off daughter cells, just like you would see in a normal retina or in the development of the cortex. A little bit later on, at day 146, you start to see this lamination, where you start to see different cell types coming, finding their way within the structure. You do lose interretinal structure, so the ganglion cells that were born early, they've got nothing to project to, and they're also limited by diffusion from the media outside of it. So they tend to die off a little bit, and we're working on ways in which to maintain that, and we think we've got something, but my interest is almost entirely in the outer retina, so this is a nice thing for me because it pushes everything towards the cell types that I'm more interested in. But you can see out here, this is ML-Obson, red and green-Obson, so you can see the nubbins of outer segments, and then here's the outer nuclear layer where the photoreceptors are, and there are some misplaced photoreceptors too, some photoreceptors that lost their way. And then ultimately at day 200, you can see not only do we see this photoreceptor layer, but the cones, which should be a single layer on the outermost portion of the outer nuclear layer. Indeed, find their way on the outer portion. Here's some outer segments that are sticking out here, but in red are all the red-green cones, and then the rods form a four to five cell layer thick nuclear layer beneath that, just like they would, and essentially this is very similar to perifobia, so not phobia, because that would be all ML cones, but not periphery, where we would see a much higher ratio of rods to cones. So if we pull this up a little bit and compare this to a similarly staged third trimester primate retina, it's pretty close. It's not, no cigar, but it's pretty close when you look at the cones and where they're arranged, and then the nuclei of the rods beneath that. And again, here are some misplaced photoreceptors as well. Interspersed in all this would also be bipolar cells, I'll show those later, and there will be some ganglion cells too, but we do get some cell loss over there. Okay, so just to show you how these things look in 3D, because it's easy to show you just little bits of, let's see here, little bits of things, little rainbows, and say, oh, that's what it all looks like. There's variation around these organines, so depending on what side of the moon you look at, you can see differences. So here in green are rods, and red are cones, and you can see there are areas where there are more cones, and areas where there are more rods, because there are waves of differentiation. But you can imagine how if you're doing an assay, and you're taking sections, and you're saying my readout is how many rods that I see, that depending upon where you make a section here, you could be led astray, right? That's why you have to know the technology, because you can get fooled by a simple little part of a rainbow in one section out of 200 different sections. And then here's, I said before, here's a marker, Vizix II, which is an early marker of progenitor cells, but in later retina it becomes a marker of bipolar cells. So here's where you can see the bipolar cells. You can start to see an outer plexiform layer, or a synaptic layer forming here. I'll get back to that. And again, rods, cones. Here's the ratio of S cones to LM cones. So that's, you should have way more LM cones than you have S cones, and that's true. So here's one S cone here, one blue cone. The rest of these here lined up in the outer portion in orange are all red green cones. And then if we blow the whole thing up so that I can show you some more structure, once again, what I'm showing you here is an area where it has mostly cones in red, and their outer segments are here, right up here. Their cell bodies are here, and then they have these long axons, and they end in a very specific cone-like snaps called a pedicle, which are very broad-based. Now there's one rod in here, I'm gonna show you in a second, it's gonna be yellow, and you're gonna see this tiny little outer segment, and then you have to follow it all the way down here to its cell body, and then it ends in a much smaller, more punctate sort of synapse called a spherial. And the red will go away here in a second, you'll be able to see that one right there. So, and here are your displaced ones. And interestingly, they express all the right markers in the right order, but they just don't look as pretty. So I've been telling you these are outer segments, do we know they're outer segments? Yes, our lab and other grad groups have shown that, so if you, in an ideal world, what we'd see are these nice stacked discs, but that's really in a mature photoreceptor. And remember that we're taking things really in the stages embryologically, so we've got kind of a mid to late third trimester outer segment, and so in that case, what we expect to see is something a little more disorganized. So we see the discs, they're not stacked, they're a little haphazard, they're starting to get there, and they're not really pretty. You see this connecting cilium here, so it has all the different elements, especially at the base of the connecting cilium, but it really is kind of a late field stage. Here's a close-up of what those discs look like. And on the other end, the other business end of a photoreceptor is the synapse. I alluded to that a little bit earlier, and here's that synaptic layer, and rib eye is a synapse marker, and the photoreceptors and bipolar cells make a very specialized synapse called a ribbon synapse found there in cochlear hair cells and bipolar cells and photoreceptors. So it really is specific in the retina for a photoreceptor synapse, and so if we look, what we're looking for are these ribbons with vesicles around them filled with glutamate that are gonna go into this cleft that surrounds either a triacid, absolutely horizontal, and bipolar cells. And sure enough, on EM, that's what we see. We see these electron micromicroscopic dense structures, these ribbons, and people up front can see all the vesicles that are surrounding it, and they're surrounding the dendrites of either a bipolar cell or a horizontal cell. There's actually another one down here, and they respond to light, and this is work that we're doing with the roundic synop. And not only do they respond to light, but they respond in a wavelength-directed manner, so we can patch clamp onto cones and determine whether it's a green cone or a red cone based upon its spectrum of response. It's graded and repetitive, so we can go back and increase the intensity of the light and see greater and greater responses. So that's really exciting to us so that we can start to look at different model systems and look at the function of the organoids and not just the histology and ICC. Okay, so that's a lot of background, but I thought I'd do that prior to talking to you about a couple of applications of the technology so that at least I could try to convince you of what the system can do, but also what it can't do. So it's not perfect, as I've shown you today. So you have to right-size the questions that you're gonna ask to the system that you're using always, whether it's an animal model or, in this case, a culture-based model. Now, in the afternoon, I'm gonna talk about modeling and drug and gene therapy testing with our best disease work, but today what I wanna touch on are some cell-based therapeutic work that we're doing as well as using this technology to come up with new genetic diagnoses. So we can convert all of the entire protocol to GMP, which means that we can do it in a manner in which we can use it for patients. We can pull out those photoreceptors so we can get essentially a purified population of photoreceptor precursors. And this is just an example here. We're lucky to have a GMP facility two floors down from my lab at the Weissman Center of Wisconsin. So we did the early work there. We subsequently converted it over to opposite therapeutics. Now, the simplest thing to do would be to be able to take the cells and place them in the sub-retinal space without much fanfare. But as I've shown you, these are very specialized cells and they have an orientation. They know what's up, they know what's down. And while if you just put them into the sub-retinal space, they may be able to figure things out or it may just be probability, and some of them will and some of them won't. And so that's gonna limit your response, that and about 1,000 other things. But if we're able to seed them perhaps into a scaffold, then we might be able to hedge the bets in the favor of having them have the correct orientation. And so we published last year at Advanced Materials the work that we did with a couple of fantastic engineers, one's a biomaterials engineer, that's Sarah Gong and Jack Ma is a microfabricator. And he works mostly on microchips for computers and he's world famous for that. And I said, well, can you do the same sort of minute manipulation for a biologic like a photoreceptor that you can for a computer chip? And he loved that idea. So sometimes you dangle a new carrot in front of somebody and get them really excited. So the initial design, this is kind of what it looks like here. There's a degradable and a non-biodegradable form. We use always the biodegradable. It can be made for pennies on the dollar. The initial mold takes quite a bit is expensive, but once you make the mold, you can make over and over again these sheets for pennies. And if you look up close, they have what we call a wine glass designs. The idea was to capture an individual or maybe two photoreceptors and then have a through channel that would hopefully mechanically get them to send their axons down that so it would be oriented. We check modules to make sure it had the right stiffness so that it wouldn't be too stiff for the retina or too soft to manipulate in the sub-retinal space. And then we use, this is actually the reason why they're red is we engineered one of our lines to express TD tomato in all the photoreceptors. And so that allowed us to, without staining them, be able to do live imaging. And this is one of these scaffold, a portion of one of these scaffolds seated with photoreceptor cells. And you can see that they have, at least they're sending something down those through channels. And this is poking out the other end and this turns out to be the end of an axon. And so if you do staining, you see that SNAP-de-Fyzen, which is a pre-synaptic photoreceptor marker, they're clustered here at the very bottom. Here's another example of that. So being able to orient them in these structures is important. The problem with this was that the payload was so small. So we were able to get most of these wine glass wells filled with photoreceptor cells but it took a lot of cells to do it and ultimately you had your ratio of biomaterial to biologic was very high. And so we wanted to limit that. So we switched to this new second generation design which is more of an ice cube tray design. So each one of these then has nine through holes within it. And with that we can have a much larger payload. We essentially quadrupled the number of cells we can deliver and reduced the number of the amount of biomaterial inversely. So one quarter of the amount of biomaterial. So the amount that has to biodegrade in the subrenil space is much, much less. And so when we do this in a rat, this is a S334 turrinopsin degenerative rat model, amount of Matt Laval's models. And we use this because it's severe degeneration. So by the time we even look at these animals they have a couple of cones but essentially their outer nuclear layer and photoreceptors are entirely gone. So we don't have to worry about biomaterial. We're also doing a xenograph with human into rat so we can use human specific markers. So the idea of are we seeing the donor cell or are we seeing something transferred to a host cell is obviated. So here's where we see the human nuclear marker here. We've recapitulated or reformed an outer nuclear layer. In the control animal, there's nothing here. This is just, this is inter nuclear layer and bipolar cells and then all the photoreceptors are gone. We use a photoreceptor marker. They're not perfectly oriented by any means. And I don't show you the control where we just do dissociated cells. But A, we get much, much better survival. We're able to control the thickness of the outer nuclear layer and the region in which we place it. So there's no gravity forces where it's kind of, there's more in one area of the bleb than in the other. And there is an orientation that's much better than if you just do dissociation. And there's an interaction. This is a little bit of just abstract art, I think for a lot of folks, but here's the cells, the human nuclei here. It's kind of a blow up here. And then we're looking at a marker for the host bipolar cell processes. So in a situation where you've got a totally degenerate outer nuclear layer. There's nothing for the cells to integrate into because there's no outer nuclear layer there to begin with. So you're trying to recapitulate that layer, not have the cells migrate up into the bipolar cell layer or further. So in that case, the idea or the concept of integration really is process integration. Are the axons of the photoreceptors and the dendrites of the bipolar cells and the horizontal cells intermingling to lock those two into place? And are the synapses being co-expressed? And the answer to that is yes. And we're now doing rabies virus tracing to be able to show that these are hopefully are functional synapses. And we started to do this in large animals now. And this is with Juan Amaral at NEI and Kapil Bharti's group. This is a pig. And so this is, I'll kind of walk it through here. Here's him making the sub-retinoblab in a wild type pig. This is really just kind of handling it and not really looking at function because these are normal pigs. So we wanted to be able to see, can we get these into the proper space? And we're using a lot of the same techniques that Kapil is using for his scaffold and RPE studies. There's a lot of stuff up front, but you'll get more interested in here in a second. So here's the injector. And you can't see it because it's transparent, but here's the biodegradable scaffold. And that's filled with photoreceptors. So it's a photoreceptor patch. And then here's him placing it into the sub-retinal space. This is the first pig. And he was the first to tell me these are kind of sticky and not super easy to move around. And so we're taking that feedback. I'm not a retina surgeon, I'm a pediatric ophthalmologist. So we're taking, here's an intraoperative OCT showing he gets into the right space but with more manipulation than we would like. Granted, these would be totally blind individuals, but we'd like it to be a little more gentle than that. But that was a first pass. And so we're doing a lot of these in pig and canines at U of N. Okay, so that was talking a little bit about cell-based therapeutics and how I do it on time. Okay, all right. So the next thing I wanna talk about is how we can use that same technology to help with folks who have retinitis pigmentosa but don't know their gene defect. And so these days with advancement and genetic testing, if somebody de novo walks into my clinic about two thirds of the time, if we just send them to various testing centers, we can get a diagnosis. We'll understand, be able to tell the parents what the gene defect is. And that's more and more important as not only just the RP-65 and Spark Therapeutics product, looks turn it, is now FDA approved, but there are so many other gene therapy and other clinical trials that are coming online. And so you wanna know if your kid or yourself are eligible for these. And you also, it's peace of mind to know that people, who in the world is working on this, my particular disease? So this is a very distressing part of the pie to be part of. Now, interestingly, our organoids make, and some of these are RPE based, but our organoids make all of these genes. So essentially they express all the genes that are known to cause retinitis pigmentosa or inherited outer retinol degenerative disease. And so our thought was that, okay, so if with whole axiom sequencing and even whole genome sequencing in the latter being just a whole commission of various SNPs, we can't find it. It's very likely that what we're missing are mutations in these known genes, but in areas in introns that are causing cryptic splice sites or other problems in the expression of the gene. So in the transcriptome, but you can't biopsy the retina, but we can make an unlimited supply of anybody's retina. So essentially we can offer the geneticists a biopsy of that patient's retina. And then you can combine the genome sequencing or the axiom sequencing, the genome sequencing with the transcriptome sequencing from the organoids and then try to find these cryptic splice sites or otherwise hidden mutations. And so we did this with, and Mike Varkas is actually working with Meg here. And so he had dinner with him yesterday, which was very nice, but also Eric Pearson and Revital on his laboratory. We've looked at a number of different families. And here's one in the Boston area where they had no diagnosis, but two affected daughters and one unaffected son. And I won't go over the clinical features of this patient, but they were known to Eric, but I was blinded to that. And so he had done a whole genome sequencing on the blood of all three of these individuals and had come up with an idea, but it didn't explain the disease. And so we had a grant together. And so what we did is made iPS cells from each one of these individuals, created organoids that I look a little different. And that's because I purposefully made them so that they had a good amount of RPE in them. Cause we didn't know whether it was an RPE-based defect or a neuro or a photoreceptor-based defect. So we wanted to hedge our bets and submit samples from all of those. And so we've purposely picked those that had more of these tufts of RPE on them than other ones that might, we might want to be more pure neural retina. We then made sure that they were equivalent to one another in terms of the amount of cones versus rods versus an S cones versus LM cones, cause we didn't want to introduce artifact. And if we're going to be looking at genes expressed in, and perhaps levels of genes expressed in cones or rods or RPE, we wanted to make sure we had roughly the same levels or very closely the same levels of those cell types in each one of the control and pro-band and affected sibling samples. So we did that just to make sure we were all on level playing ground. Again, keep bearing in mind the limitations of the system. And then when we did that, and we kind of unblinded the whole thing, there were a number of SNPs that segregated with the two affected and not the unaffected individual. And then there were a number of splice variants that also segregated with the affected versus the unaffected. But of those, only 12 overlapped and only one had a cause of the mutation that introduced a cryptic splice site. And that was in the gene CNGV3. So this causes achromatopsia. And it turns out that was the picture this child had. The child had a picture of achromatopsia. And so what we found was that there was a single base pure mutation in an intron that introduced a cryptic splice site such that what we saw was this introduction of this vestigial exon, which caused a premature termination and non-functional protein from that allele. That's important because CNGV3 is moving towards clinical trials. So in this case, we took a family in Boston who did not know why their two daughters were blind or had severe low vision. And all of a sudden we're able to say, not only do we know what it is, but there's a clinical trial coming down the pike that they may be able to benefit from. And so that's, I think, a really important way in which we can use the technology to affect patients without having to put it back into that patient. Now we also have the benefit of having those organoids to use as a testing system thereafter. So we can test gene therapies, we can test things in the organoids that we made to do the transcriptome analysis. And so in this case, what we did is we got, since I hadn't worked with CNGV3 before, we got an antibody, we validated it in post-mortem human tissue and made sure that the CNGV3 localized just to cone outer segments, which it did. So we knew we had a good tool. Again, validating your tools is key, both the system, the culture system, as well as your antibodies and everything else. And then when we looked at the unaffected sibling, that brother, we also saw, again, not nearly as pretty as a real human retina, but we have our little outer segments here. And if we look at expression of CNGV3, we see that, again, they're co-localized in these outer segment regions. But in the proband, as well as the affected sibling, we see that not only is it not present in the outer segment, but it's mislocalized. And we see mislocalized protein throughout the cytoplasm and in the inner segments. And so this gives us a phenotype to look at if we wanna test for base editing, gene editing, gene augmentation type therapies, which we're doing. And we can test promoter strength, promoter specificity, duration of promoter expression, things like that using these systems, which otherwise would have to be done in a rodent model. And in particular with gene editing, because your GRNAs that you have the fashion have to be specific to that region of the axon that you're trying to cut, that can't be done in any other system except a human system, and actually that patients depending upon what their genetic variations are. So to finish up then, the system has a remarkable capacity to recapitulate, especially the outer portion of the human retina, but it really reaches a late third trimester stage. So we're not seeing stack disks, we've yet to be able to get this to get together with RPE, so there's definitely limitations to the system, but as we learn more about that, we'll be able to apply them with more confidence. But for right now, it's important to right size the questions that we ask so that we can get real good real answers out of them and not just a bunch of artifact and start chasing our tail. So with that I wanna thank the laboratory here, some of these folks have moved on to other jobs in industry and academia, and the collaborators that were mentioned throughout the talk and the support that we have specifically for this project, and then thank you very much. Any questions, or I will cede the floor? Yeah. Thank you for a great talk. Good to see you. Nice to see you too. So did you have to troubleshoot how quickly the biodegradable, is there a certain time? Yeah, we had no clue. Yeah, so it's PGS, which has been used throughout the body, and even in the eye too. The first thing that we did was took white rabbits and just put it in the vitreous, and so we took the most inflammatory thing we possibly could, we had the benefit of working with Tom Holman out of industry, and so he said, well, don't even go forward unless you see that it's not gonna kick up a ton of inflammation. So we did that, and it was inert, which was great, and then we could see it degrade in the vitreous, and it was pretty slow in the vitreous, which is fine. We don't want to dump a whole bunch of biomaterial in the subrenal space all at once, but it was months and months, it was a little bit, I mean, what do we want? We don't even know what we want, right? So like, okay, I don't want it to be three days because that's an awful lot to have to absorb, especially with the first generation. I also don't want it to be two years, you know, so I don't know, I just, month, you know? And so you can tune that with these, but the give and take is the modulus and the handleability, right? So if you make it too soft and too quick, then it sticks to things, and the surgeon who wants to kind of push it a little bit further underneath, it grabs the retina of it, and so these are all pieces that Sarah Gong, who's a biomaterial person, can say, oh yeah, we can layer this or layer, but of course she doesn't know the retina. So the thing she wants to layer it with, you know, blows things up, and so, you know, so it's this really interesting and very fun back and forth. The arduous part about it is the testing, because you can do all these things, but until you get it, first we did it in rats, which is fine, but that's an entirely different, you know, we're going transclerally with that, we're shoving it in there, so when you move to large animals, which you don't want to do, because it's expensive, with something that you don't know your, you don't hope to be very close with, but until we did the large animals, we're like, ah, there were issues that we didn't know about until we went in that range. But we're close, I think we're happy with that. It's not gonna be a first wave, because it's a high level of complexity to both manufacture it. We can sterilize them just fine, but it will take longer for that to get, if it ever does, it would take longer for that to get to patients. But I like the fact that we can control it, but you're also introducing a lot of variables that may not be good. So the more we can make it biologic and less synthetic, the better off I think we'll be. Well, thank you very much.