 and I think we are officially on air. Hello everybody and welcome to the first session for 2021 of the Sussex Vision Seminar Series within the Worldwide Neuroinitiative. I'm George Cafetsis, a former master's student in Thomas Soiler's lab and relatively speaking newly arrived PhD student with Tom Baden. As your host for today, I would like to first and foremost thank Team Vogels and Panos Bozellos for this very initiative towards a greener and much more accessible seminar world. And having said that, allow me of course to get back to the reason we all gathered here for today and introduce our guest from Berkeley in California, Professor Richard Kramer. His bachelor studies in biology at State University of New York were followed by a PhD in neurobiology at Berkeley during which time he investigated pathways and channels of pacemaker neurons in aplasia. Later on and throughout his career, his work has come to encompass secretory cells olfactory and retinal neurons and of course cortical or hippocampal cells with a focus always on understanding and possibly controlling ion channel function. Following post doctoral years in Brandeis and Columbia universities, he then moved in 1993, if I remember correctly to Miami as an assistant professor. And eventually in 2001, he returned to Berkeley where he has been located ever since. And nowadays is C.H. and Annie Litcher in molecular biology of diseases and director of UC Berkeley NIH Vision Science Corps. Recipient of many awards for his biomedical research and co-author with James Chambers of the book, Photosensitive Molecules for Controlling Biological Function. It is with great pleasure that I'm leaving the stage for him, Professor Richard Kramer for a talk entitled Understanding how photoreceptor degeneration alters retinal signaling and how to intervene to rescue vision. So without any further ado from my side, please all welcome Professor Kramer. The stage is all yours. Okay, great. Well, George, thank you very much. I think that you and Tom Bodden and other people who have put this together and doing a great public service, having these talks that can be seen all over the world. And I hope that this extends beyond the pandemic and becomes kind of a way of life because interacting with people even if it has to be remotely is really important. And I think this is keeping things going and I really appreciate your efforts. So I wanna make sure I'm sharing my screen. Well, let's see. Is this, whoops, that isn't working now. It's swapped. We are watching the... Oops, okay, sorry, yes. Okay, so I am going to talk... Sorry, Professor Kramer, we are still watching the presenters mode. We're watching the presenter view. Play from start, now. Still, if you go to the bottom, to the, sorry, higher left corner of your screen. Yeah, what's the problem here? They didn't have two versions of this open. Second here, let's see. How about now? Still, but in this window, if you look at the upper left corner, it says swap displays, if you try that. Okay, yeah, I'll do it again. Yes. Now you're seeing the right one. Okay, great. At least the opposite, before. Okay, well, so I'm gonna talk to you today about our strategies for improving vision. And I'm gonna, you've already heard probably about other technologies that people are developing for trying to put light responses back into retinas even after the photoreceptors have regenerated. And we're gonna be using somewhat of a different strategy that depends on what we've learned about how the retina changes during the process of degenerative blindness and how changes in neurons and changes in the circuitry can be exploited to try to bring back vision when it's being lost. Not when the photoreceptors are completely gone but as they're deteriorating. And so you'll hear more about this in a minute. So, okay, so there are diseases of the retina that cause blindness by eliminating rods and cones. And the most severe of these diseases is a retinitis pigmentosa, an inherited disorder where the rods specifically have mutations that cause degeneration, eliminating the, why is this not working? Eliminating the ability of the retina to respond to light. The rest of the neurons in the retina survive but as we will be seeing change some of their properties over time. Most importantly, the retinal ganglion cells retain their normal connectivity to the brain. And so over time, the photoreceptors become completely lost and the loss of photoreceptors which starts out just in the peripheral part of the retina can extend into the central retina completely leaving people without light perception. So RP is an inherited disorder. It's relatively rare affecting maybe one out of 5,000 people in the Western world. Much more prevalent is age-related macular degeneration which is a disease of aging which affects something like 10 to 15% of people in their mid-80s. Instead of the photoreceptors in the periphery dying it's the photoreceptors right in the center of the visual field that die off. And this is only a small part of the retina but unfortunately it's the most important part because it's responsible for high acuity vision. So people with AMD have impaired ability to recognize faces to read words on a page. And this is a huge public health problem across the world. There are treatments that can stave off the progression of AMD, particularly some of the sequelae sort of effects that make the problem worse involving the vasculature. But the fundamental process of rods and cones dying is irreversible. They do not normally regenerate. And so once you've lost these photoreceptors your vision is permanently impaired. And so over the past decade or so, technologies are arising to try to restore responses not to the photoreceptors that are gone but to the downstream neurons in the retina that still survive and somehow make them respond either directly or indirectly to light. And so with the hope of course, being that you can restore people's visual perception once again. And so these technologies are several new technologies that have emerged. So this is a diagram showing the progression of photoreceptor degeneration. When people lose all their photoreceptors at end stage degeneration, then you can really start thinking about using artificial means to photo stimulate or electrically stimulate their remaining retinal neurons. And so people are developing optoelectronic devices, retinal implants that are now clinically in use in humans with severe blindness. Of course, optogenetics like channel rotopsin two and its variants can be transduced into neurons. Stem cells can end up differentiating into newborn photoreceptors. And the hope is someday we'll get them to integrate into the remaining neural circuitry and restore vision. And then finally, I'm gonna tell you a little bit about this last method, which we call photo switches. And these are small photo isomerizable molecules that might be useful as drugs for restoring light perception. All of these sort of tools are really supplanting whatever they're really designed for end stage degeneration. They sort of act as a substitute method for allowing the retina to become sensitive to light. And so that and because of sort of safety reasons, they're not really being considered to be used yet on people whose vision is just progressively declined but not quite there yet. And so I'm gonna tell you today about another approach that comes from our studies on these photo switches really that is maybe applicable to a much larger population of patients where vision is impaired but not yet completely gone. But first I'll tell you about photo switches and how they work. So this is work that we've been doing for many years in my lab and it started with a collaboration with Dirk Trauner, who's now at NYU. Dirk was a chemist and Dirk and I together with some really excellent students and postdocs in our labs developed a class of molecules called azobenzene photo switches. And this is an example of one of these. Azobenzene has this double bond, this sort of high energy double bond between these two nitrogen groups. And this molecule can exist in either the trans configuration as shown here or the cis configuration as shown here. And wavelengths of light very close to or in this case right in the visible spectrum are capable of photoisomerizing the transform to the cis form. And the idea is to take ordinary chemical groups like this quaternary ammonium and stick them on the end of the azobenzene and have those chemical groups do their work for you. So quaternary ammonium, the other name for it is tetraethylammonium and it's a common blocker of the poor of voltage gated ion channels, a whole variety of voltage gated ion channels. And so these molecules can act from the inside or the outside of voltage gated channels. But if you put them on cells and then you wash them away as what would happen if you introduced them into an organism, the only way they're gonna remain around is if they somehow get concentrated and trapped inside of cells. And it just so happens that even though this is a fairly large molecule, there are pores found in many types of cells that are capable of conducting molecules like this, allowing them to concentrate in the cytoplasm where they can interact with voltage gated channels. And their interaction with these channels is dependent on whether they're in the trans or cis form. And so essentially you can block and unblock voltage gated channels by switching the wavelength of light. And it also, so, you know, like I said, the channels that are large enough to conduct these molecules are quite specialized and not found on all cells. One such channel is trip B1, which is found on nociceptor neurons. Another channel is the P2X receptor. And it turns out completely by accident that, so this is an example of what happens if you put this molecule, this azobenzene photo switch on a neuron that has trip B1 channels and you activated trip B1 channels, you now gain control of action potential firing in that neuron because you're blocking and unblocking its channels. The molecule is essentially acting like a light sensitive local anesthetic. And so you've basically controlling the firing of neurons without any genetic manipulation. So this is optogenetics without the genetics, if you will. So in the context of the retina, it just so happens that during degeneration that there are changes that occur in retinal neurons that make those neurons permeate permeant to these kinds of molecules. And so, you know, this is sort of just an outline of how we typically do studies on the retina. This is one approach we use a lot which involves using a multi-electrode array and using a mouse strain that has the equivalent of retinitis pigmentosa. So it quickly loses its rods and cones even as they're developing. We isolate the retina from one of these animals, put it on a multi-electrode array and record spiking in response to light or just spontaneously. And here you can see a wild type retina responding at light onset, very robustly. This, you could see to some extent, this retina has cells that fire at light offset and you can reliably get responses from this isolated retina mediated by the rods and cones that are still healthy in this normal retina. But now if we do the same experiment on a retina taken from an RD1 mouse, you can see that there is no light response. We can turn on the lights and turn them back off again. And all of these neurons in this raster plot which is recording from, you know, 50 neurons at once in this array. You can see none of them are really responding to light. Every once in a while, we'll see one out of a hundred cells appears to be an intrinsically photosensitive retinal ganglion cell that does respond to light. But they're in a very small minority. For the most part, there's not much of a light response. If that, however, we take this same retina and we soak it in one of our photo switches and I think this one was called D-NAC and then we wash away what's in the bath leaving any molecule inside the cell to do its thing. We see that now these neurons respond very vigorously to light. And the light response is quite large. It subsides somewhat over time. We think that that is not due to the photo switch itself relaxing, but rather due to intrinsic mechanisms in the retinal ganglion cells that cause spike frequency adaptation. And so there's no sort of decay in the ability of the retina to respond over many, many cycles of light and dark. We could do this for hours as long as the retina stays alive in a dish. So we've managed electrophysiologically to take this retina and at least some of the neurons are now responsive to light in a sustained way. So I wanna emphasize this is not number one responding normally. The response is not as rapid as it would be in a bile type retina. All the cells respond with the same polarity of light. They're all firing in response to light. There's no off cells in this. If there are off cells, they're responding like on cells in this situation. And we don't think it's all the neurons responding to light. It's just a subset. And I'll show you some, let's see, move ahead of myself here. Movies here, I'm so far ahead of myself. There we go. This is a movie I'm gonna show you of a retina expressing G-Camp, the genetically encoded calcium indicator. And you will see, and this is a blind retina. You'll have to take my word for that, but that's been treated with a photo switch called BNAC. And you'll see every time the light goes on, a subset of those neurons will respond. It's not all of them. The retina is sitting on top of the multi-electrode array. So you can see those black dots and the leads leading to them. Those are the electrodes themselves. And it's something like a quarter of the cells are responding to light. And I'm just gonna tell you that we've been able to identify in a rough manner of speaking, which neurons those are. They are, they have their dendrites in the off sub lamina of the interplexiform layer. So they are off center retinal ganglion cells. So even though they're responding in a positive manner to light, they're actually the off RGCs that are responding. And this is a loop. So you're seeing the same responses repeating themselves. Okay, so I'll go back to what I just skipped. So I'm not gonna talk too much about the photo switches per se. I really wanna, I wanna talk about them and tell you where they've led us because there's some surprising findings that have come that have been revealed just by seeing these photo switches at work. So just to sort of tell you where things stand about whether BNAC or its relatives might be suitable for restoring light responses not only in mice and rats, but in humans of course we're interested in that. And some of the properties of BNAC are really suitable for human use as a possible vision restoring drug. So the wavelength you need to get light responses is safe. It's in the middle of the visible spectrum. This is the Lambda max is about 480 nanometers. So that's safe. You need bright light to get retinal neurons to respond. But the light is, it's in a photopic range, but it's not exceeding the maximal level of light that is short of causing damage to the retina. But it does mean if someone is gonna use this system they would need to have some under ordinary lighting conditions or indoor lighting conditions. They'd need to have some image intensification hardware like special goggles in order to see the world. We've been able to show that various kinds of visual behaviors can be restored in blind mice simply by injecting them with BNAC. And we've done a pretty extensive set of studies looking at toxicity in mouse and rat by compounds like BNAC and they don't seem to be very toxic at concentrations where you can reanimate these blind retinas. And we've been working on sort of a related problem which is delivering enough of the molecule to work without it being toxic but having it stay in the retina or stay in the eye for a long enough period of time to be useful. There's a practical limit to how often you can inject somebody with these drugs. So we've been working on delivery systems that will allow the molecule to be released continuously over a long period of time which I don't have much time to talk about. So, okay. So that's all I'm gonna say about that possibility which we're very excited about and I'm happy to talk more about it about photos which is really being themselves therapeutic tools. There's a lot involved in taking a drug candidate to market and some of our work along these lines is being passed off to pharmaceutical companies that are much better than we are at doing that sort of thing. But it's, you know, I'm still very interested and we're all hopeful that it's gonna work and represent a valid approach for restoring some degree of vision for certain people with types of blindness. Now along the way, I really wanna tell you about something completely unexpected that came from studying these photo switches that started out as a puzzle but now it's opened up a whole new field really with a whole set of its own possible opportunities that might be therapeutically relevant. And so the initial observation that led to this is gonna be shown in this slide. So a student in my lab several years ago named Yvonne Toczitski who spearheaded a lot of this work was studying these blind retina and he really wanted to know where VNAC worked. I mean, we could see the sort of rise of light responses in retinal ganglion cells in blind retina but that response might be intrinsic to the ganglion cells or perhaps it's referred onto the ganglion cells by some other upstream neuron in the retina. So a simple way to discriminate between these possibilities was just to whip up a mix of neurotransmitter receptor antagonists that block all the known neurotransmitter receptors in the retina at least by rapid neurotransmitters. So that's glutamate, GABA, glycine, acetylcholine even dopamine receptors and put them on this blind retina and now ask if we put BNAC on the retina or DNAC which is a similar molecule are we still able to make the ganglion cells light sensitive? And the answer was a very resounding yes. If anything there, the light responsiveness is even more robust than if these receptors were not blocked and we confirmed that we were really blocking synaptic transmission with this set of blockers. We could, you know, in wild type retina we could see absolutely no light response get through. And so this result tells us that yes the retinal ganglion cells themselves are targets for this photo switch molecule and their intrinsic properties are being affected by the photo switch. And so then just for the heck of it Yvonne decided to do the same experiment with a wild type retina that still had rods and cones but, you know, because the receptors are blocked the same cocktail of blockers is used the rods and cones are pretty much irrelevant because their light driven signal can't make it to the ganglion cells. And so the question is does DNAC or BNAC is it still capable of photosensitizing the ganglion cells? And we fully expected the answer would be yes because we're far downstream from where changes are happening, which is the death of the photoreceptors but in actuality you can see here that the light response is completely absent. So not only aren't rods and cones able to drive the ganglion cells but DNAC is unable to confer light sensitivity. And that implies that something is dramatically changed in these retinal ganglion cells during blindness. And this was kind of a new realization, you know people had noticed that over time there are anatomical changes called remodeling that happened in the retina when rods and cones died but the fundamental process of synapse transmission of, you know, intrinsic activity most other things were thought to be pretty much the same in a degenerating retina than in a wild type retina. And in fact, you know, retaining the connectivity is essential for any of those technologies to effectively carry a signal from this degenerated retina to the brain. If you wanted to introduce stem cells into the retina and have hope that they're gonna restore somebody's vision there needs to be, you know pretty reliable unaltered transmission of information from retina to brain. But this kind of result suggests that something big is happening in retinal ganglion cells that people hadn't been aware of before. And so we looked into this further. I'm gonna skip these slides. So it turns out one of the big things that happens is, and this is relevant to the photo switches is that the membrane permeability of retinal ganglion cells increases dramatically. And one way to show this is using very common fluorescent dyes. So here are images of retina flat mounts. You're looking at the retinal ganglion cell layer. Here, the retina has been stained with a nuclear dye, nuclear ID, which just labels all cells and binds to DNAs, so all cells look red, including vascular cells and retinal ganglion cells. And then it's counter stained with Fitsi, Fluorescine isothiocyanate which is an incredibly commonly used dye, green dye, which ordinarily, if you just put it on a retina, it doesn't stain anything because it doesn't cross membranes very easily. But if you put it on the retina of an RD1 mouse, lo and behold, the ganglion cells pick up this dye. And you can see they're ganglion cells because you can see their cell bodies and their axons. But it's not every ganglion cell, it's just a subset. And it turns out if you block P2X receptors, which are the receptors that I said are changing, they're getting upregulated in degeneration, then you don't see this Fitsi label. So we've shown this with several different dyes. We confirmed that it's really the P2X receptors that are changing. This doesn't happen in a knockout mouse that's missing P2X receptors. So the P2X receptors are upregulated and they're becoming hyperactive in these degenerating retina. The effect of that on physiology is not so clear, but there are effects on physiology and degenerative retina. And many people have shown these effects over the years. And here's an example that we've collected sort of recently. If you just do a recording from a wild type retina and compare it to a recording from an RD1 retina, the level of spontaneous firing just in darkness is something like eight to 10 fold higher in a degenerating RD1 retina. And all of this extra firing, and again, it's not in all retinal ganglion cells. Some of the ganglion cells are responsible for this hyperactivity. And in fact, if you look at over time, the loss of light responses in this retina, which is of course directly related to the loss of rods and cones and the rise of spontaneous firing, there's a very tight correlation between the two. Spontaneous hyperactivity goes up as the retina degenerates. And that's only one of the problem, one of the changes that happens in the degenerating retina. There are other changes that we think are somewhat independent and that have also been noted by a lot of people. So here's a paper by Menzler and Zeck from 2011 showing that retinal ganglion cells start to fire not only more, but also rhythmically in these bursts. And the bursts tend to be synchronous across retinal ganglion cells. So there's correlated firing in retinal ganglion cells. And one way this correlated firing can come about is if a common presynaptic neuron is driving simultaneously several retinal ganglion cells and there's strong evidence that presynaptic amocrine cells are also able to have synchronous oscillations of membrane potential that are necessary for aligning the firing of these retinal ganglion cells. And this is partly due to a change in the gap junctional coupling between the amocrine cell network and between retinal ganglion cells due to the upregulation of connections. So there are several different phenotypic changes that happen in the retina as degeneration occurs. But there's one sort of question, I guess sort of a unifying question that really hasn't been answered. And that is, what is it about the death of rods and cones that initiates these changes in the first place? So, and so, sorry, I'm gonna give you the motivation for addressing that question first. This is a little bit out of order. So just this just sort of summarizes in a healthy retina, we have photoreceptors responding to light, they're transmitting information eventually to retinal ganglion cells, even if those ganglion cells are somewhat spontaneously active, the light response can be seen over the background of whatever spontaneous activity there is. In a de-generating retina, if spontaneous activity is very high, now it becomes difficult to resolve the light elicited response against all this background. And if we sort of knew what the signal was that was causing this hyperactivity, perhaps we can go back to a situation where the background was low and now sort of allow the light response to emerge once again. So it's not obscured by the de-generate, by the hyperactivity. So this is the sort of, like I said, the motivation for studying all of this, I don't know why this is happening automatically. So the question we have is, what is the signal that is responsible for initiating degeneration? So as rods die, something is somehow, the downstream neurons in the retina know that the photoreceptors are dead and they begin to change accordingly. So one possibility that a few people have revealed in recent years, including Felice Dunn and Daniel Kirstensteiner is that when you lose photoreceptors, and this is common throughout the nervous system, when you lose excitatory drive from neurons, there can be changes in downstream circuitry that try to compensate for that loss of input. And that's known as homeostatic plasticity. And both Felice and Daniel have demonstrated very nicely examples of homeostatic plasticity in the retina involving individual types of photoreceptors that are lost. So perhaps the hyperactivity you see in the retina is simply a consequence of homeostatic changes due to changes in excitatory drive, let's say. That's a possibility. It might contribute to why there's hyperactivity, but we don't think it's the primary explanation and I'll show you why in a few minutes. But I also just tell you the punchline here what we do think is mostly responsible for this. And we think it's retinoic acid. And so what is retinoic acid? Retinoic acid is a downstream metabolite of retinaldehyde. So in the retina, vitamin A, otherwise known as retinol, is picked up by the retinal pigment epithelium, which enzymatically converts it into retinaldehyde. And retinaldehyde is released in very large amounts from the RPE. It's supplying all of the opcins in rods and cones. And that retinaldehyde is normally absorbed and covalently attached to opcins. And the photoreceptors act as a huge sink for that retinaldehyde. And when the photoreceptors die, it's possible. We don't know exactly why, but there is an excess of retinaldehyde. We do know that when there is an excess of retinaldehyde, it can be converted into retinalic acid by this enzyme, aldehyde dehydrogenase, which is actually a big family of enzymes. The specific isoforms of LDH and the retina are known as retinaldehyde dehydrogenases. And they're expressed throughout the retina in different cell types. And they produce retinoic acid, which is a really important signal in early development, both in the eye and throughout the body. Normally in adulthood, there's not a lot of retinalic acid around and its whole signaling system tends to be shut down in adult cells, particularly in adult nervous system. The way retinalic acid works is by crossing membranes and binding to a protein called the retinalic acid receptor, RAR, which dimerizes with another protein called RXR. So this complex of retinalic acid and its protein targets bind to very specific sites in DNA called retinalic acid response elements and turn on genes. And so this is extremely well studied. It's a very big part of developmental biology. This whole system has been an important target for cancer research. And there's been a big effort to develop drugs to interfere with the system in an attempt to treat cancer. But like I said, in the adult nervous system, this system has largely been thought to be shut off because there isn't much retinalic acid around the adult. So, but we think there's retinalic, we sort of tested the idea that retinalic acid is what's causing these phenotypic changes in degenerating retina. And I should give credit to the people who inspired this idea and they were Robert Mark and Brian Jones at the University of Utah who had shown retinalic acid was involved in some of the morphological changes. And so we decided to look at the physiological changes. So, what would constitute evidence that retinalic acid is a signal for plasticity, for physiological plasticity. Well, first of all, it would be nice to be able to see that there really is higher retinalic acid. Retinalic acid itself is hard to see. It's very labile, breaks down easily. But instead of seeing it directly, you can see the consequences of retinalic acid by making a reporter gene where the retinalic acid response element drives expression of GFP. And just as a control, we have a constitutive promoter driving expression of another fluorescent protein, RFP. And so if we package this thing in a virus and then we inject this in the eye and infect neurons with that virus, you can see in wild-type mice, you get the expression of the RFP, which is constitutively expressed, but you don't see any GFP expression. Oops. If you do this in an animal, this in this case, it's a rat with retinitis pigmentosa. This is a mutation that's a truncation of rhodopsin, causes photoreceptor degeneration. You can see now that not only do the neurons express the red fluorescent protein, but also the green fluorescent protein, indicating higher retinalic acid. And we've done this same experiment with RD mice. So there really does seem to be higher retinalic acid signaling. The next sort of question, if there really is, if retinalic acid is really the trigger for these events, you should be able to mimic these events by exogenously adding retinalic acid. And, oops, this shows photo switching. This is, we're back to using one of our azobenzine photo switches. I said that they work really well in conferring light sensitivity on degenerating retina. So here's an RD-1 retina. This is a photo switch that responds to different wavelengths of light, the technical reasons we were using this photo switch. And you can see that this blind retina is now responding to light. And here is a wild type retina, not responding to light when exposed to the same photo switch. But if we inject retinalic acid, this is all trans retinalic acid itself into the eye of these RD-1, of this wild type mice and wait five days, we see that it starts to behave very similarly to the degenerated retina. So we've been able to mimic the phenotype of sensitivity to photo switches in a wild type retina simply by adding retinalic acid. So we can see retinalic acid, we can mimic retinalic acid's effects, but perhaps most importantly, we can block retinalic acid's effects on hyperactivity by injecting a blocker, I don't know why this is happening automatically, by blocking, I don't see an easy way to stop it. So if we inject a blocker of a small molecule called BMS493, which blocks the retinalic acid receptor, so it blocks the signal transduction triggered by retinalic acid, we see a reduction in the hyperactive firing of retinal ganglion cells, and we see this happen significantly. And once again, it takes several days after the injection to see this effect. And that's because what you're really interfering with is a transcriptional event that takes several days to exert itself. And if you wait too long after injecting this chemical, then the effect wears off again. So this is evidence that the hyperactivity can be not only prevented, but actually reversed by blocking the retinalic acid receptor. Okay, so it's not just the sort of hyperactivity. I mentioned that the retina shows these very characteristic oscillations of activity. Here is a whole cell patch recording from a retinal ganglion cell. This is an alpha retinal ganglion cell. And you can see, if we hold the membrane potential of this cell at zero millivolts, you could see these oscillating synaptic currents, probably coming from amicron cells, representing inhibitory postsynaptic events needed by GABA or glycine. And they occur at something like seven to 10 Hertz. They're very characteristic of degenerating retina. And you don't see them in wild type retina. And if we take an RD1 mouse and inject saline in one eye and inject this drug BMS493 into the other eye, these oscillations are completely absent from the recordings from the retinalic acid blocked, retinalic acid receptor blocked eye. And this is sort of spectral analysis of these synaptic currents. You can see that there's a big peak around seven Hertz, which is a little bit faster than what other people have described. This classic 10 Hertz oscillation, we're doing our experiments at somewhat lower temperatures, I think, which may account for the somewhat different frequency. But after BMS, these oscillations are gone. And this is group data from many cells from several different animals. And this was work done by a visiting professor in my lab, Scott Nowie, who deserves a lot of credit for doing this. And I should also mention, just because mentioned more than once, Michael Tilius is really the brains behind a lot of this retinalic acid work and a lot more of what's coming up next. So, if we put together the data from the past few slides, I think we have evidence that retinalic acid is present when these sort of phenotypic changes are happening in the retina. It's necessary for the changes to occur and it's sufficient for the changes to occur. And so, we feel that the data is pretty strong, that it is the initiator of a lot of these, at least a lot of the phenotypic changes, there may still be role for other processes like homeostatic plasticity as well. So, okay, so now that we know that retinalic acid is central to this, we can start to ask what is this really due to vision in an animal where the photoreceptors are not yet completely lost. So, for this, we turn to a somewhat different strain of mouse called the RD-10 mouse that degenerates a little bit more slowly than the RD-1 mouse. So, when a two month old RD-10 mouse, the sort of normal light responses mediated by rods and cones are still present. And you can see in response to a period of light, the ganglion cells, some of them fire vigorously outlight onset, some of them fire vigorously outlight offset, some of them fire at light onset and light offset. And this looks pretty much like a normal wild type response. After a month of age, the light response is really starting to diminish dramatically. And that's because the rods and cones are dying. And then after two months of age, the light response is completely gone. So, in front of our eyes, these retinas are degenerating. And if you, sorry, if you look at this carefully, you'll see that as the photoreceptor, as the light response is diminishing, the level of background firing is increasing, just as I've described before. And here is this graph again, correlating the two. Now, the idea is that retinolc acid is involved in this process. So, if we intervene somewhere in this process, for example, midway and block the retinolc acid receptor, question is what will happen to the animal's vision really? What will happen to the ability to pick out light responses against this rising background of spontaneous activity if we block the initiator of this hyperactivity, retinolc acid. So, we decided to do that. And here's an experiment where we once again took an RD-10 mouse and right at that moment of time where we're halfway through degeneration. Sorry, this is really bothering me. I'm not sure how to get it to stop automatically advancing. But right, one eye is injected just with vehicle. The contralateral eye is injected with this retinolc acid, with this RAR inhibitor, and we wait five days. And if you look carefully here, you can see at this point in time, this 50 millisecond light flash doesn't produce any noticeable light response in this MEA recording. But if you look in this eye, very carefully right at the light flash, there seems to be some extra spikes right after the light flash. And if you repeat nine light flashes and average the response, you can clearly see the light response emerging from the background, which is diminished also by the blockade of the retinolc acid receptor. And so this is recordings from five vision-impaired mice comparing one eye injected with vehicle and the contralateral eye injected with this retinolc acid response, this RAR inhibitor, retinolc acid receptor inhibitor. So you really can augment light sensitivity, you can augment light responses simply by reducing RA signaling in these mice. And so that's one way to block the retinolc acid receptor. There's another way to block the retinolc acid receptor and that's through genetic means. So it turns out that there are mutations in RAR that act as dominant negative mutations. This is a gene product that not only doesn't work in terms of activating gene transcription, but it actually eliminates the ability of wild type RAR from turning on genes. So it's a dominant negative mutation. We can encapsulate that mutant in an AAV virus injected into the eye, targeted to retinal ganglion cells and now in a very genetically targeted way prevent retinolc acid signaling in retinal ganglion cells. We've done that and looked at light elicited behaviors in mice. So here is an assay where we're asking mice we're pairing a flash of light with a shock and mice very quickly learn to associate light with this very bad outcome. And so they freeze in response to seeing a flash of light and they do so if once they've learned this at a very dim level of light, they begin to show this light avoidance behavior and an RD10 mouse where it's photoreceptors are degenerating don't show this behavior until you get to higher intensities of light and maybe even then they don't show it at all. But if we inject this RARDN virus, now the RD10 mice behave indiscriminate from normal wild type mice. So we've been able to restore this light avoidance behavior. Okay. But we've wanted to do a little more sophisticated visual behavioral tests involving spatial vision, so we've been able to assess sort of the perception of light in general. And so we've devised this assay that was inspired by Ed Pugh who's at UC Davis nowadays, where we have an operant conditioning task where we have mice exposed to a pattern a drifting grading on a screen and we make available a reward to get their nose into this little nose poke device. So the sequence of events is the stimulus comes up. The reward is made available. The animal has some short period of time to go get its reward. We iteratively run this shortening and shortening the period of reward making the task worn more stringent. And eventually the mice learn they better get the reward when the stimulus is actually there. And then once the animals learn this, we can vary the contrast of the stimulus and see how well the animals can recognize these stripes as the image gets more and more ambiguous. And what we see is that mice learn to make this association very quickly when they first start to train. They are randomly going to get their reward, but over one night of training, they get better and better at doing this. They jump off their running wheel and get the reward very short latency. They retain this behavior, even though their photoreceptors are degenerating for some period. But if you look at the same mice late in degeneration, they start to fail miserably. And that's because they can't see the, they can't discriminate the stripes in this image anymore. And so if we vary contrast, sort of partway through degeneration, we can actually, we can determine where the threshold is for the mice being able to see this stimulus. And you can see here, this is too low of a contrast. The mice are failing miserably. This is a pretty good contrast. At a hundred percent contrast, the mice get a right all the time. And as you look at mice over the time course of degeneration, you know, they have one contrast sensitivity early in degeneration, but late in degeneration, they're pretty poor. And even the highest, even full contrast is not generating a full response. And you can compare changes in their contrast sensitivity in this manner. And so the idea is to intervene in this retinolic acid pathway and see if we can improve their contrast sensitivity by doing such interventions. And I've already told you about one intervention, which is this blocker of the retinolic acid receptor, BMS 493, which is a chemical made in a laboratory. It's never really been a drug. It's a drug candidate, I suppose. But there's another molecule we can use in this pathway, which is an FDA approved drug in the United States and elsewhere. And it's known as disulfurum. And I'm not going to go through the whole story, but disulfurum is a drug that's given to hopeless alcoholics. It deters their desire to drink alcohol because it causes very bad hangover like symptoms if they happen to have a drink. Its common name is called anti-abuse. Maybe some of you have heard of it. It blocks the synthesis. It blocks the breakdown of aldehydes, really. And in doing so, it blocks the synthesis of retinolic acid. So we have these two ways of intervening in this system that we've been using to look at, you know, spatial vision. And so here's data from both of these. This is the same kind of contrast sensitivity curves I showed you in RD-10 mice. Here's their ability to do this task early in degeneration versus late in degeneration. Here's what happens if you inject the RAR inhibitor BMS 493 in their eye. You basically take mice that were functionally blind and give them the ability to detect at least high contrast images. And then this is disulfurum, which you don't inject. Instead, you can introduce it orally by simply mixing it up with food. So here's these RD-10 mice, some of them given a disulfurum in their diet. Some of them given no disulfurum in their diet. You can see this dramatic difference of disulfurum rescuing some portion of their dramatically increasing their contrast sensitivity in this task. Just in the last couple of slides because I realize I'm running a little over. We've collaborated with a group at UC Santa Barbara. This is Michael Gord, an old friend of mine, actually UC Berkeley alumnus. He was one of my TAs or one of my neurobiology classes years ago. And he does cortical calcium imaging from visual cortex. So these are cortical neurons expressing GCaMP and they're color coded based on their orientation selectivity. So you can expose the mouse to different orientation of bars of light and dark and define the orientation sensitivity of these visual neurons. And he's doing this in RD-10 mice. And what he finds is that the orientation tuning of cells and cortex is greatly heightened in these mice undergoing degeneration if you inject BMS493 in their eyes. This is the tuning of one, two particular cells, one from a treated animal, one from a control animal. And this is sort of an average of their orientation selectivity index. It gets dramatically better when you block retinoc acid signaling. And he's also doing experiments now with disulfurum and doing experiments with natural scenes showing that neurons respond more reliably to natural scenes when you block retinoc acid signaling. So the sort of improved vision carries over to cortical responses and behavioral responses in these vision-impaired animals. I guess this is really the final data slide. So in retinitis pigmentosa, people lose their photoreceptors across the whole retina. But in macular degeneration, people lose their photoreceptors just in a small little area of the retina. Of course, over the macula, mice and rats don't have a macula. They have a phobia. They have a rod-dominated retina. So you can't really directly ask questions relevant to AMD and a mouse. But you can do a poor man's simulation of this by wiping out the photoreceptors in sort of a small island of the retina. And we've done this in collaboration with Daniel Polonker at Stanford by putting in a little piece of foreign material under the retina. And we've done this in collaboration with the U.S. and the U.S. with the retinal implant or even a piece of aluminum foil, aluminum foil, if you will. And the consequence of that is that the photoreceptors adjacent to that material die. So it kind of simulates local photoreceptor degeneration. And what we see is local activation of hyperactivity in the ganglion cells overlying that area, whereas the rest of the retina looks sort of normal. So we think that the changes that happen globally across the whole retina when photoreceptors die are occurring in microcosm in very local areas when just a small group of photoreceptors die. And that leads to the possibility that maybe these changes are occurring in AMD, but the difficult thing to directly address since AMD is pretty specific to human beings, even other primates, there are not really great other primate models of AMD. And so just sort of to summarize, we have this major public health problem of people losing photoreceptors. Most people, a large population of people lose photoreceptors either to RP or AMD without losing all of their vision. They just become vision impaired. They experience what's known as low vision. And a small fraction of those people go on to have no light perception whatsoever. We have a variety of, you know, really heroic methods for restoring restoring vision to people with no light perception. But yeah, and that, you know, there's been a lot of effort to advance these technologies, but you know, there's been a lot less effort to address this low vision population that still has some photoreceptors left that people would be very reluctant to risk with one of these sort of heroic methods that may supplant or even eliminate the remaining photoreceptor. Elicited responses. So we think that, you know, inhibitors of retinal gas, signaling or perhaps synthesis, either drugs or genes may represent a way to improve vision in these people. And, you know, these two methodologies are not mutually exclusive. You can imagine them being used in complementary matters to, you know, improve the other methods that are possibly restoring vision. So I really, this is the most important slide and I was going to start with this, but I want to make sure I have enough time to mention the people who did the work. On the left here are the people currently in the lab or just recently left the lab who did most of the retinolc acid related work. Kevin Kau is doing a lot of two-photon imaging of retina. He's a current postdoc in the lab with a background in chemistry. Bristol Tenlinger is a veteran of the lab. She's moving on to a position in biotech. She's done a lot of work on RD10 mice. Zach Health has graduated with a PhD in vision science. He did a lot of the early work and he's an entrepreneur starting a new company now. Scott, as I mentioned, is a long friend, an old friend of mine who is, it's really fortunate that he's in the lab as a visiting professor for a few years. Michael Tilius is currently on the job market and as I said, he's the sort of brains of the outfit who has done a lot of the retinolc acid work. And we have collaborators, the ones that were featured in this work were Daniel Polonker and Michael Ward and people working in their labs. And we've been supported by a range of different philanthropies and the National Institute of Government Agencies. And years ago we started a small company that's puttering around with the idea of commercializing photo switches for potential commercial use. So I'm happy to answer as many questions as people wish to ask. And thank you very much for allowing me to go a little over time. That is great. Thank you very much for the talk. And for sure, don't worry about slightly exceeding the typical time window. I mean, the meeting is virtual, so if people are interested and they are indeed, they can stay for longer. So before I start moderating the post-talk discussion session, I would like to remind the audience that they can either ask their question in the chat, in the YouTube chat or join us in 10 or 15 minutes from now in the Zoom room that we are sitting for a follow-up, more informal conversation if you wish. Now it looks like your effort was coordinated. You were here giving the talk and Michael Telias was already addressing some of the questions. But nevertheless, I will make sure that all of them reach you in case you want to add something in them. The first one is from Justin Grassmeyer. And you partially also addressed it towards the end of your talk. But he says, hello all. Happy Martin Luther King Junior Day. Has similar work been done in cone dominant or modular containing retina? Any reason to suspect these findings would be different in AMD or cone degenerations? I don't know if you want to add something to what you already said. I mean, there hardly has it, you know, that's a very good question. And there have not been really, as far as I know, there haven't been really great for cone dominant retinas with the equivalent of RP. To some extent, people have been using rabbits, which I'm not sure that's really cone dominant, but there's more cones, certainly. Mutations in rod specific genes are most often the things that lead to retinitis pigmentosa. And so if you want to use a line of animals that reliably have these kinds of issues, you're sort of almost by definition going to be impairing function of rods. But then, of course, can end up affecting cones and how dying rods sort of lead to cone death has been studied a lot. There's factors like rod-derived cone viability factor. I mean, I would say that there are people I know that are trying to simulate some of this by using laser ablation of cones in the fovea of primates. It's a much more difficult system. It's sort of, you know, everything goes a lot more slowly with primates. It's really quite a convenient thing to have a line of animals that are undergoing a very reproducible, predictable course of retinal degeneration. So there's a whole lot less work on condominated retinas. Yeah, I mean, in the models you have, like even the stereotypic progression of the disease is complicated by itself to begin with. I should say, of course, there are dog models of RP. And in dogs, you see the same hyperactivity. Whether retinoc acid is involved is not yet established in dogs, but the phenotype seems quite similar. The next question is from Vladimir Kefalov. And I think it's a really good question. Like I also wanted to ask you that if the excess retinoic acid in the generating retinas originate in the RPE, does blocking out RPE65 to block retinal production also block the structural and functional changes in the inner retina? Right. So I should say that that is only our working hypothesis that it's originating on RPE. We don't really know where the retinoic acid, that we know it's, if we inject retinaldehyde itself into the eye, we can mimic these changes. If we block, like I said, RALDH, we can prevent these changes. So we know that retinoic acid is, I'm sorry, the retinaldehyde is the source of the changes. Retinaldehyde can be produced in mule or glial cells as well as in RPE. We can't really say which cell, cellular source, you know, cell type is the source of RPE. It just sort of makes some sense to me that it's likely to be RPE. We haven't done the experiment that, that Vladimir is suggesting, but it might be a good one. Down that road, like one question I have is, like you mentioned what happens to wild type retinas when you inject retinoic acid, like you saw what happens with the firing rate of RGCs. And my question is, what happens to the photoreceptors themselves? Do they degenerate when you just inject? We actually had to do a control like that and they do not degenerate. That was, that would have been a confound in that experiment most definitely. Manoj Kulkarni asks if this effect of retinoic acid signaling, if it is uniform across all RGCs or only limited to off RGCs or other subpopulations. Yeah, that's a good question. That we, we have not looked at, you know, been a been so we, that's one thing we'd like to know. We think that the hyperactivity is maybe selective for off RGCs. That doesn't mean that retinoic acid doesn't have other phenotype, you know, cause other changes in other cell types. You know, basically every cell in your body has retinoic acid receptors and is capable of transducing higher retinoic acid into some change. We also have some evidence that the bipolar cells are changing their properties. And just about every cell type we've looked at, exhibits some change. And so, you know, this is one of the reasons that this is a very fertile, you know, there's a lot of fertile future possibilities for work. There may be many different things going on in different cells that and they remain to be seen. Thank you very much for that. The last question on the chat at least, and I'm already starting to allow people in the zoom room. So in case you want to, I posted the last question on the chat is from our very own Tom Baden. And I think it's a really interesting question. So, and if I'm allowed, I will add a comment at the end of it. So he writes, so the fact that retinoic acid signaling increases naturally during the generation, can that be linked to it doing something else, which is actually useful, which you are now interfering with. And the comment I would like to add on that, like we briefly discussed it on a previous meeting we had, is that also when you injure the optic nerve, like you see that some developmentally active and then silenced path, transcription paths, are again reactivated. So upon injury, the developmental, some developmental stages are re-initiated. And like maybe retinoic acid that has a role in the retina development reasons, when the retina degenerates, it is somehow signaling, you know, a reversion to that path. Yeah, I mean, there's an old saying in pharmacology that every drug has two actions, the one you know about and the one you don't know about. I suppose that's true of any biological manipulation. So whether retinoic acid induced gene expression is having some beneficial effect that we don't, we are not aware of yet. Obviously I can't say anything about that. But you know, there's other questions like how far downstream, you know, so bipolar cells are changing, retinal ganglion cells are changing. What about at the other end of the retinal ganglion cell in the brain? Are synapses, you know, in lateral geniculate changing are the behavior of those LGM neurons changing and as cortex changing. And so there may be, you know, constructive changes sort of propagated along the system. But, you know, it's not always, there are many examples in the nervous system where plasticity isn't always constructive. One example is epilepsy. You know, in response to traumatic brain injury, you know, people often emerge from injury feeling perfectly fine for a while, but then, you know, days, weeks, months later develop epilepsy and because of plasticity. So, you know, I don't know whether there is some adaptive reason for this system or whether it's just maladaptive. Yeah, I see your point. So that was officially, if I'm not mistaken. Yeah, the last question that appeared in the YouTube chat, the link remains there. So if people would like to join us in this Zoom room, feel free to. This transmission will terminate in about five minutes, maybe less, so make sure you join in time. I would like once again to thank you, Professor Kramer, for your talk. It was really fascinating, also including the unpublished results like of the behavioral and the cortical experiments and just to remind to the audience of this series, of the Sussex Vision series that next Monday we will be hosting Professor Simon Laughlin back to the Zoom room that we are currently seated in. As you can see, more and more people are joining. So like I wanted to ask you, like maybe my question is too naive, but do we have simpler models, like not in mice, I mean, maybe zebrafish, for example, of such diseases? And again, this question is inspired from the regeneration following optic nerve injury that we have some models that do regenerate the erections and some that do not. I don't know if you really consider a zebrafish retina more simple than a mammalian retina. I don't know. There are certainly degeneration models in zebrafish. I don't know if they're sort of slow, slow onset the way that for the TANR. And none of this has been looked at yet in zebrafish, but there may be, you know, certainly the genetics is simple and there may be good reasons for looking at zebrafish. I'm not sure simplicity is one of them. Yeah, I'm just inquiring because like I have been studying the optic nerve in the past and according to literature at least zebrafish and the simpler evolutionary speaking organisms can regenerate the erections following injury, but mice cannot. So, you know, maybe it's something, you know, worth of a look at least. Just to make sure everyone is following us like now I'm not officially the moderator of the discussion. It's supposed to be informal. So please, if you won't go ahead and ask any question or make any comments, you might have. Well, hello Noga. It's great to see you. Absolutely. She must be, I think. Maybe she's still following us from the YouTube. Transmission, which is always 15 seconds delayed because of the feedback, but that will stop soon. So let's see how do I. Yeah, probably that was it like 15 seconds later. I don't know if she heard my group. Yeah. You are muted unfortunately. Maybe I can try to unmute here myself. Let me see. Maybe she doesn't have a microphone connected. Yeah, because I cannot find a microphone for sure. Took service. I don't know why. No, no, it's okay now. I hear you. Good to see you. It's been a very long time. I know, I know. Yeah. Even beyond the pandemic unfortunately. But anyway, I was just saying that I really enjoyed the talk. And I haven't made any question that I might have missed from your talk. So you recorded an improved license for us after you blocked. So. After you blocked there. So Noga, could you please switch off the YouTube video and just follow us. Within the zoom room. So we don't have the feedback. Okay, I can make it easier by also switching off the live transmission. Did it help.