 All right. I'd like to welcome my friend, Samar Hazzar, today. Samar is going to talk to us about, we actually didn't get a chance to talk about, what he was going to talk a little about. But Samar's career has basically been looking at non-photoc, or non-visual inputs to the visual system, looking at circadian rhythm biology, how you set the circadian rhythm biology through retinal ganglion cells. He comes to us, I did his postdoc at King Waiow's lab, and then got a position at Johns Hopkins, and now he's become the director of circadian rhythms at NIH. So it's been kind of a fun career to watch. He first, I first sort of became aware of your work, was your first talk at ARBO in 2001 or 2002. I had a poster, and I wasn't sure I believed what he was saying, that we could actually get enough integrated signal in two leaflets of a ganglion cell membrane to actually communicate the information and set the circadian rhythm. But since then, his work has dramatically expanded, and he's really informed us as to how we set the circadian rhythm biology. So it's very much a pleasure for me to post Samar here, and please. Thank you so much, Brian. Thank you everyone for coming so early, and it's really a pleasure to be here talking to people who understand the retinal, which is something that I usually have a hard time presenting because we work on so many other areas. So most of the stuff here is gonna be like a lab meeting. So I would really love your input on the stuff. So I have four stories that I'm gonna tell you about, and I'm gonna go through introduction pretty fast. I really agree with Brian. It took me a while to believe the data myself, to be honest, because it's just really surprising, and hopefully we'll have a nice discussion about the evolution of these photoreceptors and how they are integrated in the retinal. So I really appreciate you stopping me. I think it would be nice for this to be informal because I'm presenting stuff that I usually don't present in my talk. So that's why I have the title so general, non-roton, non-con-retinal photoreceptors if you want some vision in general. And by vision, sorry, before I go, I would just like to acknowledge the people in the lab that helped me do the work and actually do all the work, to be honest. This is my lab at Johns Hopkins University. These are the people still there. Actually, it pretty just graduated, and she's in Houston. She just went there before the major hurricane. We were part of the mouse trial lab. Myself, Hype, Chazelle, and Regis Gravilla, three mouse geneticists that actually showed reagents, resources, and lab members, which allows us to discuss areas that we usually don't think about. And I really suggest this for any young faculty that is starting to try to team up with another young faculty because then you have bigger input, bigger lab meetings, and this allows you to enjoy the science. And this is my body lab at the National Institute of Mental Health. Karine is a new person coming in October and I just hired three more people that should be added to this list. Okay, so as I told you, I made the title really general about vision because we all know the function of the retina for image formation, but the eye detect light for functions that are subconscious, that we're not aware of them consciously, which include the construction of our pupil that although very simple, requires a very complicated path through the olivary protective nucleus all the way to the ciliary muscles. And circadian photoentrainment, which happens in the suprachiasmatic nucleus, these nuclei that contain the biological clock just above the optic chiasm. So to understand the retinal function or the eye function, you really have to look at the totality of the function of the eyes. And the way we study this in our lab, we use a very simple system. It's actually incredibly robust and simple. This is really experiment. This is not, my students when I used to present to undergraduate, they say, oh, this is a nice animation. I'm like, no, this is really experiment. This is actually a behavior. So here you look at the mouse for 32 days here and then 40 days here and you change the light, dark environment. So open areas are light, gray areas are dark. And what you see here is we're running activity. I'm also running on the wheel. And just to put it in perspective, this small peak here is approximately 300 we're running revolutions per 10 minutes. So if you calculate the amount of free revolution and make it in a human terms, the mice run approximately four to six miles every night if we're running on a treadmill. What you could see slowly from this is that if the mice can detect the light, dark environment, they have a 24 hour period exactly matching the day and they can find their activity to the dark portion of the cycle because they are nocturnal. But what was really remarkable and what got me into the scientific field is that even if you put these animals into constant dark conditions, what you could see there is a still beautiful clock. It's not exactly 24 hours, hence the name circadian. In the case of mice, it's less than 24 hours. It's 23.6 hours. If I put you in these conditions, you should be going in this direction if you have a normal circadian clock because the human circadian clock is 24.2 hours. So what you could see is that even without any environmental input, there's an endogenous, molecularly determined clock that allows the mice to measure with quite accuracy the onset of their activity without any environmental input. Clearly you could see this would be evolutionary and disastrous because now the mice would be active in the middle of the day and that would be very bad for them. They are easily picked up and eliminated from the gene. So the importance of the biological clock is to confine your activity to imported niches in the light, dark environment. We could actually jet lag these animals. So if you bring them from the United States to Europe, for example, New York to Amsterdam, you could see that it takes them six days to adjust to a six hour shift, just like us. You go to Europe for six days. By the time you're fully adjusted, you come back to supper again in the return trip. So what's really remarkable about this and the behaviors in vision is that compared to other behaviors which have good behaviors, but this is incredibly robust. You look at this, you could actually estimate with great accuracy the circadian clock, you could estimate the photo entrainment and you could estimate the speed of entrainment and if you change the light intensity and the color of the light, you could figure out which are the best light information that is required for entrainment. And we'll talk about this later. So I don't have to explain to you the retina. People have thought only rods and cones are the photoreceptors in the retina, but there was a paper, two papers back to back with negative data in science. That's how surprising it was. And this is when I was starting to look at postdoc where the authors from Russell Foster Group produced animals that completely lack rods and cones and hence image blind. Yet these animals beautifully photo entrainment and I have to admit this animal is from my lab, their actograms didn't look very good to explain it, but so this is a rod-less, coneless animal from my lab, but you could see it beautifully photo entrain to the light-dark cycle. In fact, pretty similar to the wild-tag animals. So this indicated there are other photoreceptors in the retina. This was heretical, as Brian said. A lot of people didn't believe it. I presented these papers and I tried to find holes in them and I couldn't in the journal club when I was in the University of Houston. So to make a long story short, it was really, really the pioneering work of David Burson, Ignacio Provincio, and then in collaboration with us when I was in King Waiyal's lab that now we know that there is a subset of ganglion cells. I think these are the ones. Kahal, Yundan, I'll talk about this in a second. A subset of ganglion cells that are intrinsically photosensitive, in other ways they are photoreceptors, because they express the melanopsin, which is similar to cone-opsin and rhodopsin in the rods and cones. And what's really interesting that if you knock out rod, cone, and melanopsin, which take away the intrinsic photosensitivity of these retinal ganglion cells, now you have an animal that completely frees on throughout the light-dark cycle. So now even though this animal forms an eye, connects to the brain, cannot photo-entrain to the light-dark environment. In fact, in a subset of blind humans, blind humans that are both blind for image and non-image functions, these patients always grumble about cyclical problem in sleep. And you could see here beautifully why it's cyclical. If there's sleep, if you imagine this is a sleep onset of the patient, if the sleep onset falls at the right time of the social day, then the patient have no problem going to sleep. But what happens to the patient is that they start having jet lag in their own town, as I call it, where now their onset of sleep is completely out from the solar cycle, from the solar social environment. So now we know that after so many years of research on the retina, that there is a subset of ganglion cells when we first discovered them, we thought there's only 700 or 1% in the mouse retina that are intrinsically photosensitive. And what's interesting about, sorry, what's interesting about these ganglion cells is that they could respond intrinsically through the melanopsin system. But even if you remove the melanopsin protein, they could still act as ganglion cells by incorporating rod and cone inputs. So there is nothing about them that differentiate them from other conventional ganglion cells. In fact, awesome retinal biologists have classified these as ganglion cells and gave them name because they receive rod and cone inputs similar to other ganglion cells. So once we saw that they are ganglion cells, I was lucky to have Randy Reed in the same department at Hopkins and Randy said, you know, in the olfactory field, because olfactory neurons direct connectly to the olfactory bulb. People use the tau-lacty system to label the cell bodies and axons of the olfactory neuron to trace them to the olfactory bulb. So I said, oh, we could do this for melanopsin. They are ganglion cells. They are projection neurons. Let's just replace them with an olfactory gene with the tau-lacty marker gene, labeling these cells blue after you same with X-gans and labeling their axons which are all coursing toward the optic disk. And what was really remarkable, even though we predicted it, is just seeing the brain was just mind-boggling, is to see that these ganglion cells project specifically to the suprachiasmatic nucleus, the circadian center that I told you about. And in fact, in the field, there has been another circadian center that people talked about, known as the entogeniculate leaflet or IGL. This is close to the image-forming centers in the brain, the DLGN, the dorsal latogenical nucleus. And what you could see, these fibers come and innervate this leaflet and avoid the image-forming centers in the brain. So clearly, we get so excited that we call them circadian photoreceptors. Lasted two weeks, and then we found that they project to the pupil center. So clearly, they are not only circadian photoreceptors because they seem to also project to the pupil region. In fact, when I started my lab, I did a very extensive determination of where these fibers go. And in addition to the circadian and pupil centers, we were amazed to see that they project to sleep and alertness centers. That's why you're alerted when you get bright light and you feel sleepy when people dim the room too much or use black background on their slides. I tell people, never do that. And mood-regulating centers. And if you come to my talk today, I'm going to tell you about a very exciting region that we discovered that maybe it's the depression center in the brain, but that's for the later talk. So when we first discovered IPRGC, we thought they only constitute one subtype, which we called the M1, but really work of three amazing women in my lab, Jen, Tara, and Tiffany. Tiffany actually has her own lab now at Northwestern University. Have found that IPRGCs are diverse. There are at least five subtypes that can be easily differentiated based on a very specific morphological and electrophysiological properties and Melanopsin expression. What is really amazing for us, and Tiffany is continuing this and she has a story, hopefully it's going to come out soon, that these M4 cells are actually your typical on-alpha ganglion cells that are important for contrast detection. That was quite surprising to us, because as I told you that M1 cells avoid image-forming centers, but here's a subset of IPRGCs that project to image-forming centers. And I'll be happy to discuss this later if you're interested. But the diversity of IPRGCs gave us a dilemma. How can we study the individual contribution of each subtype? I'll be happy to tell you how we achieved this, but again, Jen, that you saw in the previous slide, and Ali, who has his own lab at UVA and just published a beautiful current biology paper recently, they actually created an animal where, I'll tell you the details later, where they only removed the M1 population without removing the other non-M1, and that's because the level of expression of Melanopsin is highest in this population. And in these animals, what was very important, even though we killed nearly 700 M1 IPRGCs, vision, as measured by the optomotor vision test, was completely normal between wild type and these animals that lack the M1 population. But now we ask the question, what happens to non-image-forming visual functions like the pupil constriction? Now, pupil constriction is mostly dependent on rod and cone input, so we expected it to be relatively normal. But we actually were stunned. So here's the wild type mouse. If you put it in the dark, the pupil fully open, this is infrared light. You give moonlight level, the mouse pupil is very sensitive. It really constricts more than 50%. If you now give a high light, it constricts a very small level. Now notice that the animals that only lacking the M1 IPRGCs still have normal vision, normal rod-cone input, constrict to level at high light intensity that is less than low light intensity. So in fact, we get 1000 times less sensitivity in the pupil just by deleting 700 cells. So this excited us a lot and we thought even though the pupils are more open, maybe circadian photoentrainment would be affected a little bit. But again, it was stunning, the data were stunning. Here is circadian photoentrainment of a visually capable mouse where you cannot differentiate these animals from the triple knockout animal that I showed you. So even though the animals can have conscious image formation if you can't believe in consciousness in mice, they can actually free run throughout the light, dark cycle. So when Ali came to me with this result, I said, Ali, this is just really mind-boggling. This is like a mathematician that can solve the most complicated equation but she cannot add one plus one equals to two. That just doesn't make sense. So he said, let's do something that the mice despise and people used to stress the mice, put them in constant light condition. And what you see in wild type animals, they start running much less on the wheel and they lengthen their period. They don't have the motivation to jump on the wheel when the light is on. These animals completely don't care. In fact, the amount of running on the wheel is not different, the period length is not different. So clearly then, there is an evolutionary separation between image forming function through the majority of RGCs and non-image forming functions through the M1 IPRGCs. And interesting, that separation doesn't happen in the brain, it actually happens in the retina. So again, another evidence that our retina is part of the brain and it always drives me crazy when National Institute of Health say, oh, you work on the retina, not on the brain. I just get a little bit angry with something like that. So this was really exciting, but what we found that there is even specialization in these M1 IPRGCs, and I'll show you the data in a second. So you guys know more than me that different streams are sent to the brain of different characteristics of the visual field and it's assembled in the visual cortex. So we thought for a simple non-image function that detect the presence or absence of light, why would you need that? You don't need specialization. All you need is their light or is their darkness. But this work by Melissa Ciment who has her own lab now at Oakwood College and Anna Chen who has his own lab at National Taiwan University, they decided because of several reasons which I'll be happy to discuss, to make melanopsin Cree line with the Brain 3B LOXP stop DTA line. So if you make these animals only cells where the two promoters are active will end up expressing the DTA. So you're only gonna kill cells that both express melanopsin and express the Brain 3B transcription factor. And when Alan did this and looked at the M1 population, we were kind of stunned because the SCN is beautifully innervated but the OPN as well as other areas that I showed you were completely innervated. This gave us a very simple prediction that these animals should have a defective pupil constriction but normal circadian photo entrainment. And in fact that's what we found. These animals had the same defect in pupil constriction as the animal that lacked all the M1 IPRGCs. But when we looked at circadian photo entrainment it was completely indistinguishable from wild type animals. And what was really amazing is that we found that all non-M1 cells the majority of M1 cells that project to the OPN are Brain 3B positive and only 200 M1 cells are Brain 3B negative but they are exclusively projecting to the supracasmatic nucleus. So when we found this, we found this diversity in IPRGCs we wondered what do the IPRGC do more? And in data I'm not gonna show you we actually found that this population in addition to the axon that they send to the brain they also send an intra-axonal innervation in the retina. So they send an intra-retinal axon. So I said, are they doing something in the retina? So to answer this question I have to tell you about the story that kind of true have done in the lab. A very talented graduate student who actually left science looking at me how much I struggled in grants which breaks my heart and it's just a warning that we're losing really good scientist because less academic science she's working for a company. Because she looked at the suffering that scientists have to do to get money to do beautiful experiment. And just a warning for this country I guess. What she found is that we always thought that the biological clock as I told you is internally produced by a molecular mechanism. But she looked at the literature and she found some papers that are really interesting. Animals that are born without eyes, math five and my five knockout animals or if you remove the eyes at P zero the clock is lengthened similar to what I showed you in the LL experiment. But if you manipulate at P 60 the clock doesn't change. So she said maybe IPRGCs also play a role in regulating the clock development. I said Kylie, I'm a circadian biologist. Triple knockout animals had a normal clock. It just doesn't fit. She said let me just do the experiment. So she collaborated with a graduate student in the lab David McLean. And now instead of killing only M1 IPRGCs they use the very strong toxin the diphtheria toxin A subunit to kill all IPRGCs. And just to show you the difference between the attenuated diphtheria toxin which only kills M1 IPRGCs versus the strong diphtheria toxin. Here's actually 14 days after birth this toxin is being expressed. And what you could see the STN is still beautifully innervated. In fact, because of Kylie's work we know that ADTA only kills M1 cells in the adult stages. It requires a long time to even kill the M1 cells. But this strong DTA doesn't allow the STN to get innervated at all. Because it kills the IPRGCs quickly. And it kills most of the IPRGCs including non-M1. So just to remind you if you have wild type animals they photo-entrain they have a normal clock. If you have melanopsin knockout animal they photo-entrain they have a normal clock because rods and cone can signal through the IPRGCs. But now what happened if you looked at the ADTA animals which I showed you before these free run with a normal circadian period. But now the strong DTA free run with a longer circadian period. And what's really interesting this is exactly similar to what you see if you cut the eyes at P0 versus P60. So we were very excited about this we thought maybe IPRGCs provide trophic support to the STN but we looked at the STN it looked normal. So Kylie did an experiment that I thought like it's never gonna work well. She just reeled animal in constant darkness. Remember we measure circadian clock in constant darkness in the circadian field. But before melanopsin people have thought that light detection in mice happen at P12. But now with melanopsin we know that even light detection can happen embryonically as Richard Lang showed in his beautiful vasculature paper. So now she reared the animals even before birth in darkness. So she reared the pregnant mother in darkness. And if you do that not 100% of the animal but nearly 60% of the animal produce a very clear length in circadian period. We know that this is not developmentally changing the system because if you expose them to light later in development in adult stages they actually photo-entrain similar to normally raised animals. But here's the kicker. If you now put these animals back in constant darkness the period that was longer than 24 hours all of a sudden corrects to less than 24 hours. And in fact here's the data. You could see there is a very significant effect of light input that can change the properties of the circadian clock. So is that dependent on the SCN projecting IPRGCs? Yes, actually we published this before in the paper I showed you before because these animals have the normal circadian clock that you see in the wild type animals. So now we know that these M1-B negative IPRGCs not only are important for photo-entrainment but they are also important for the development of the circadian clock and the properties of the circadian clock. So then we wanted to do our control knowing that if we delete M1-IPRGCs we should not have any effect on vision. And luckily for me Kiley did two different colors of coratoxin because we wouldn't have been able to see this. So this is the contra-intervention from ganglion cells in the vision center and this is the SCN you could see beautiful separation. This has been published many times in the mouse. But now if you look at the animals where we killed most IPRGCs this segregation of ipsi and contralateral is completely affected. And here's the quantification of the data. So we were kind of shocked but we remembered that as I told you Tiffany had found that M4 cells are part of the visual system. So we wondered if, and by the way this segregation deficit leads to problem in vision. So, and it's through the retinal waves I don't have time to talk about but I'll be happy to discuss it. So we wondered if these cells are sufficient for normal segregation or whether these cells are required. So now we use the brain 3B animals which lack all these populations but maintain this simple population here. And we were stunned because the segregation in these animals is completely similar to the wild type animals. And here's the quantification in fact vision is the same in these animals. So that tells us that these M1 brain 3B negative population not only project to the brain to affect the circadian clock both properties and photo entrainment but they also project inside the retina to affect the retinal waves. But here's the coolest part. Their function in the retina is light independent. So their function on the retina is simply a relay for the retinal waves that happen in the retina. So we think and we don't have yet a very strong evidence for that. I'll be happy to get any ideas how to do this is that retinal waves going through milanopsular cells somehow through this inter-retinal axon limit the spread of the retinal waves allowing more localized waves to occur so you don't get interaction between the two eyes so you get separation of the fibers. I'll be happy if anybody has any way for us to test that because that would be the next ultimate experiment. What's the target? What's the target of you? Yeah, so that's a great question. So we think one of them is dopaminergic amicron cells. And so although dopaminergic amicron cells also make connection with the dendrites of IPRGCs, people have tried adenosium to try to record from IPRGCs and dopaminergic amicron cells when they are close to each others and they never get signal. So I think the reason is you could see and this is not just a schematic. Usually these are very far away from the cell body of the IPRGC. So maybe dopaminergic amicron cells signal in this direction to IPRGCs and IPRGCs goes back and signal to them. We should just image this. Yes, I would love to do that. Okay, last story. So we published this paper in Ivers and people don't notice it but it's like the data's, it'll be just amazing if we could do something. For that, that would be great. Okay, so that story just shows you that IPRGCs in addition to their projection to the brain they have an inter-retinal function that is actually important for image formation and in a light independent manner. So I don't know what more evidence someone would get that these may be the older population of photoreceptors because they seem to be involved in so many functions in the retina that they seem to be included in the path that is important for the retinal function. Yes. So actually you said that push on your eyes and give light sensation, but I wonder also if they might be stimulated. So that's interesting because if they are working through non-intrinsic light input, so you could imagine rod and comb signaling through them. Melanopsin clearly, if you activate it, will still signal through it. But because we've done, I passed through this very fast, because we've done all these retinal waves in the dark, we know that they are light independent. But they could still incorporate light input. So I'm imagining if you use pressure, maybe you activate your classical photoreceptors and that send information back to the retina through these IPRGCs. That's my prediction. Potentially, but these are kids that sometimes have total retinal detachment for more than two years. Yeah, so I think that actually is very interesting because you don't need up to nerve. This is exactly right. So this retina can be detached and this still gonna signal because it's inter-retinal axonal fibers. So it actually really fits the hypothesis that you're mentioning. So when we started seeing these ganglion cells, we said, wow, this is interesting. There's many things that I could tell you that are different than other conventional ganglion cells. But this is, to me, make them a hybrid of ganglion cells and amacrine cells because they send as a projection neurons, but they also seem to innervate the retina. So we wondered if they have a different developmental pathway. So one of my students, Justin Brody-Comet, came to my lab and he was very excited about studying the development of IPRGCs. I am not a developmental biologist and so all this work really is his and I'm just recently starting to understand it. So I'm gonna present it with the fear that I won't be able to answer all your questions to completion. And just as a review, in retina, there is a temporal time domain, cell-autonomous unidirectional transcription-regulated time domain to produce all the different players in the retina. And people have thought for many years and for good reason, as I'll tell you, a little bit math five or autonal seven is important for ganglion cells because if you knock out this gene, you don't have the optic nerve, you don't have ganglion cell layers, you have amacrine cell layers and all the other layers, but you don't have ganglion cell layers. And in fact, there are many mutations in many different organisms and including in a human that have these very interest, well, I don't know if you could say in humans, I'm not a human, I don't work on a human patient, so I think they are interesting diseases to the mutation, but I know that the patients will suffer an optic disc area problem, glaucoma susceptibility, retinal detachment, and even vasculature problem, which is gonna become exciting in a second when you see why I put all these things here. So the first thing we notice when we looked at the literature very carefully is that sometimes people take a dogma and you look at just one paper that people somehow ignore. There's a paper published in 2012 where people use math-3, math-5-3 line, labeled the cells, and then they back-filled RGCs, and they notice that there are a lot of RGCs that are back-filled but are not expressing math-5. This is actually kind of surprising because as I told you, math-5 knockout doesn't even form an optic nerve, so they lack 95% of ganglion cells. And I think people ignored that because this math-5-3 also labeled photoreceptors, so they thought maybe this 3 line is non-specific and so they just ignored this paper. But there was another interesting paper in 2010 that showed in the math-5 null mutant it's not that the cells don't form, it's that across development there's a huge cell death that occurs while the ganglion cells are being born. So luckily for me, Justin was smart and he thought, okay, let's try to do something, let's prevent cell death and see if we can produce ganglion cells in the math-5 knockout. If math-5 is required for the specification of ganglion cells, preventing cell death should not affect ganglion cells at all. It should be the same as the math-5 knockout, namely you don't have any ganglion cells. Here we're using markers for ganglion cells such as brain-3A and brain-3B. So first of all he did the control-backs one and the control and the math-5 knockout and we were quite amazed that if you do the math-5-backs double knockout, now you kind of restore ganglion cells in the retina. So clearly there are some ganglion cells that are math-5 independent. So math-5 is required for the survival of the non-math-5 but there is a specification of ganglion cells that does not require math-5. And here's the quantification of the data. So this is the non-math-5 lineage and this is the math-5 lineage. Okay, now interestingly, even though we restored the number of ganglion cells, we did not restore the optic nerve. So somehow the ganglion cells formed, they received rodent cone input, they have all the markers of ganglion cells we did sequencing, but they were not able to form the optic nerve. Why is that? We think because these ganglion cells from the math-5 knockout have a problem in axonal guidance because the optic disc doesn't form in the math-5 knockout. So these math-5 dependent ganglion cells playing a non-cell autonomous role in forming the optic disc, which is really exciting. I don't love to get any suggestions of how this could happen. And in fact, you see that you could increase the number of ganglion cells, but still the axons are just going all around in the retina. Here's one type, SMI-32, which actually labels on alpha RGCs. So then we looked at the literature and in the literature, the vasculature which actually forms in two-step ways in the retina here is beautifully seen in adult animals and P9 animals is abnormal in the math-5 knockout. So the hyroid goes into the retina, but it doesn't regress in the math-5 knockout animal. So we thought maybe restoring the ganglion cells and that's why you see patient, human patient, with problem with the vasculature, if they have a mutation in math-5 knockout. But actually restoring the math-5 number of ganglion cells did not help in the vasculature. So the ganglion cells, even if they are playing a role, they may have to play a role in collaboration as we were talking yesterday, Amy, with the optic disc itself. So then we found a very interesting paper that Moos Lap published in 2015, a really brilliant paper, where they looked at the literature and they know that brain-3B and islet-1 are downstream target of math-5 to specify ganglion cells. So what they did, they used the math-5 promoter in a conditional way to drive brain-3B and islet-1 in the absence of math-5. And doing this, actually they completely restored the ganglion cells, but they also form an optic nerve in this situation. So we used these animals, so you can see here they have a beautiful optic nerve and they have a beautiful NEA response. So I remind you that if we restored the ganglion cells without their axons normally, you do not restore the vasculature, but now if you use this animal where it doesn't have a math-5. So all these vasculature and axonal effect have to be dependent on these two genes. Now you restore the vasculature in this retina beautifully. So at least now we have, so unfortunately as I told you, this is a lab meeting, this is where we are right now. So now we have at least an idea that some components downstream of the brain-3B and islet-1 are required for the normal axonal and vasculature development of the retina, which is something now we could actually use to try to find component downstream of the system. And you know, anybody who wants any information to work on this, there's only one person working in my lab, I'm not a developmental lab, I'm never gonna go all out when Justin leaves, you could take the project with him. Let me know, I'll be happy to share any reagents, whatever you wanna do. One last important aspect, as you know, I showed you earlier, ganglion cells first are the first to develop from the neuroblast, they go to the ganglion cell layer. And what you notice is that in the math-5 knockout because cell death happens so fast, this generation of ganglion cells look as if it's dependent on math-5, but actually if you do math-5 back knockout, you could see that they are beautifully generated as you would imagine at the right time. So this is again consistent with the fact that for some reason there is a math-5 independent pathway that is important for the generation of these non-math-5 dependent ganglion cells. So just to modify the dogma in the field, we still think math-5 is very important, clearly it's important for the survival of these non-math-5, but there is another transcription factor which we still don't know that maybe leads to the expression of these downstream targets that is dependent on math-5 for survival. So without math-5, these ganglion cells die and if you prevent their death, they form ganglion cells, but they cannot reach to the brain. So we're very excited to try to somehow help these ganglion cells reach to the brain or somehow correct these vasculature and ganglion cell deficit. I have two more stories. Please feel free to ask me any questions on what I had thus far and otherwise, yes. So this is a very simple question, but with your physiology and the rest, it would be fascinating that you thought, did you know to try to help with insomnia, partially, but also with jet lag that people take exogenous malatone? So what, I mean, it's all kinds of issues that we just briefly had in the workshop, I mean, how well does it really get absorbed, what happened, what do you think happens when you take exogenous malatone? So there is- Do you think you can override the system? So there is no doubt that light is the strongest time giver to the circadian clock. I mean, it's just nothing comes close. In humans, in animals, maybe feeding, but feeding is good for peripheral organs, but quickly, actually, the central oscillator sticks over the peripheral organ clock after you change your feeding. So malatone works and it helps because malatone does have a little effect on the SCM clock, not a huge, but malatone is good in putting you to sleep. But the other problems with malatone is you could read on the tablet already that because it's a dark hormone, that sometimes people experience depression-like symptoms because- Wild dreams, people report that. Wild dreams, I never knew about wild dreams, that's interesting. So I think if you're a patient who's completely blind and who having major problem in, I think malatone is a good course, I would love to have a better course than malatone. Like something to activate whatever, the areas that are important in the brain for the shifting the clock, that would be the best. If you could shift the central clock to match the animal, that would be the absolute. What's this new drug they advertise all the time on TV? Yeah, that's very, the Vanda Company is non-24. Yeah, non-24, they advertise it all the time. So it's actually very interesting, it's an agonist for melatonin, but it also activates the serotonin receptors. So it seems to work better than melatonin and people seem to like it much more than melatonin, but honestly I still think it's not as efficient. And Ali, my post that published a paper that also dopamine from the VTA can actually feed back on the clock. So it's clearly a very complicated, and we know that for example, the Raffer nucleus send axons to the SCM, the VTA now we know send axons to the SCM, the IGL sends MPY fibers to the SCM. So it's clearly the central location with light of what is happening in the brain. That's a great question. I'm sorry, I can't answer that. I love the work you're doing here as well about all those additional ties because the price you pay for ignoring your normal circadian rhythm. Absolutely, and if you come to my other talk, you'll see that it's really remarkable how the system works. I mean, it has a tremendous impact on the pressure and the altitude. And actually in the other talk, we think we found the region, we're testing this in humans now to see if we have the same region, but we think we found the region for light-induced good mood at least in the brain. We'll see if it irons up. Fascinating. Thank you. So even for people who are blind, non-cited people, light therapy conceivably, we have photons that actually pass through our skulls. It really, I don't think it will help at all, unfortunately, because only melanopsis is only expressed in the retina. So if they're blind, but their eyes are intact and the system's intact, light is still critical for them. Yeah, in fact- If you have patients that they have that are anaphylamic and no lies at all. In fact, it's really amazing. I've yet to see a very strong paper. There are a couple of weak papers, but very strong paper saying any glaucoma patients have any sleep problems. And one thing I was talking about, Amy, I didn't want to put more projects here, but the IPRGC seems to be resistant to cell death. So if we could cut the optic nerve and the IPRGC is 50% of them, stay there for nearly ever. And we tried it for three months. So glaucoma patients don't seem to have sleep problems. And this- Because these cells- These cells are resistant to, it's so important evolutionally, they don't die. Yeah. Were there some studies suggesting that they were intrinsically or sensitive ganglion cell or intrinsically or sensitive cells of the virus? Oh my God, I had the whole section on this, but since it's my advisor, and I didn't want to fight, why don't I talk personally with you? Okay. I could tell you that I'm trying to be as political as I can, which is I'm not very good at that, but I'm learning being in the NIH that I have to be more political than being at Hopkins. I literally don't believe that the pupil in the iris does anything. So yeah, it's evolutionary there. Couple of cells express it. But they're not producing enough signal to be important. In fact, in their papers they say that, but somehow they publish nature and current biology. I don't know how. We're writing a paper right now to put this to rest. So Sam Raab also followed your career like Brian since the early 2000s, because I'm interested in these cells and their effect on life sense to me, but a phobia. Yes. Are you gonna talk about that at all? Not at all, because actually there is a person also in UCLA that does a lot of, I forgot her name. She's really good. In fact, I have a little bit of injury in my cornea, like small, I went, and it's so amazing how much I get affected by light and I feel like it's bright light. So I'm thinking my melanopsin cells are making me pain. I went to Finland, yes, Finland in May when the day wasn't yet that long and I've never wear sunglasses. I had to buy the darkest sunglasses. I couldn't open my eyes in the day anymore. So I think this whole idea that maybe IPRGCs connect to the cornea from the DRG, I kind of believe that. I think that's more credible than the Irish pupil muscle. Interesting. Both of them are surprising, but for sure the DRG is 10 times more credible than the Irish muscle. The Irish muscle is completely incredible. I'm just gonna put it out there. I just didn't like being political. Okay, so the next question we wanted to ask is we know that IPRGCs and melanopsin are important for photoentrainment, but we have a beautiful photoreceptors in the retina. We have these rods and cones that are just incredible, evolutionary, incredible cells. Why would you maintain melanopsin expression in ganglion cells if you have these beautiful photoreceptors? You would imagine that melanopsin will be, mutation melanopsin will have no effect on the animal survival and eventually you get enough mutation that some vertebrate animals will lack it. In fact, we did lack a copy of one of the melanopsin gene. For example, chicken, frogs have two melanopsin genes. We only have one. So why is it then conserved across all organisms? So to answer this question, we first wanted to know which photoreceptor contribute in the absence of melanopsin. I told you in the absence of melanopsin you get photoentrainment, but is it rods or cones? So Kauara Altinas did a very interesting experiment. She decreased the light intensity and she found that opposite to what people like me in the circadian field have thought, actually circadian photoentrainment is incredibly sensitive. It can actually entrain to 0.1 lux. So that already gave us the idea that although rods were thought to be saturated at circadian photoentrainment level, if you could entrain at 0.1 lux, maybe rods play a role. But what really shocked us, if you have animals where only cones are functional, so rods don't have function, melanopsin don't have, eye-parachutes don't have melanopsin. These animals actually do not photoentrain, even though when we do vision tests in these animals, their cone vision is completely intact. So now we said, what about rods? So now we did the same experiment. We got smarter, we made it harder for the animals because people didn't believe that they entrain at 0.1 lux. And you could see beautifully they entrain at 0.1 lux. It just takes them longer, but they actually beautifully entrain eventually. Now if you do animals that have only rods as photoreceptors, you get beautiful entrainment in these animals that is exactly similar to the wild type animals. In fact, the only time rods don't do very well is at very high light intensities, which is kind of weird because you would think the more lights you give, the better the circadian clock will be. So consistent with this, if you now use a rod knockout that is really good like the genat one instead of rod degeneration, now you see that melanopsin even is not capable of entraining at low light intensities. So this is a rod function. So it seems that rod and melanopsin are the ones responsible for circadian photorentrain. Just because I wanna go to that, I'm not gonna tell you about the circuit. So then the question is, why still this, if rods can't do it, they do it very well. Maybe they don't do it as well in high light intensities, but so what? Because you get enough light to entrain. Why is melanopsin conserved when these cells can receive a beautiful rod and cone input that gives them higher sensitivity to light and faster response? And they fire exactly the same in the presence of, in the absence of melanopsin or the presence, at least in vitro. So like before me, Melissa came up with a very smart experiment. See, I actually looked at the brain 3B deleted animal. Remember, I told you these animals were completely normal at photo entrainment. One thing I omitted, they are normal at photo entrainment when they have the melanopsin gene intact. Now if she knocks melanopsin, you could see clearly that these animals are not normal anymore. It takes them so long to entrain to a very simple shift in the light-dark environment. What's even more interesting, if you give them a summer day, these animals completely fall apart. And when I first presented this to the circadian biology field, they said you showed your best, especially this one, you said you showed your best actograms, so I had to put all the actograms. Here's all the actograms from brain 3B animals that are under 12 light-dark cycle. And what you could see clearly is that they don't photo entrain well at all. So even though they have innervation in the SCN, if you don't have melanopsin, there is a major problem in rod and cone compensating for the absence of melanopsin. But one could say this is an animal that lack all IPRGCs. What happens in melanopsin knockout? I told you, me and 30 other lab tested melanopsin knockout always on 12-12, and we get this beautiful effect, no effect in photo entrainment. Now if you just change the light environment, if you just lengthen the amount of light, now melanopsin knockout falls apart. So our hypothesis right now is that melanopsin is not actually a circadian photopigment at all, because rods and cones can photo entrain. Melanopsin is a photoperiodic signal. So what is the hardest function for animals and plants to actually measure? They don't have calendars, they don't have iPhones, it's the change of season. So the best way you could measure the change of season is to have a photopigment in plants that measures the day length, or in animals to have a photopigment that is not encumbered by how much light previously you were exposed to. Rods and cones are highly encumbered by the light history, because depending on how bright or how dim the environment is, rods can become either super sensitive or completely saturated. So that's why I think it's the simple reason why melanopsin is conserved across evolution. In summer days, rods and cones cannot compensate for this and animals would completely fall apart and not be able to deal with seasonality. And in fact, animals that will depend on breeding for seasonality will have a problem even breeding. And in fact, we just finally, in collaboration with Richard Lang, just developed a melanopsin knockout hamster animal. And we're gonna see if these animals now can actually mate when you give them along the light cycle. So they mate as heterozygous, but we wanna see if they can mate as homozygous. So I'm gonna stop here and I'll be happy to take any questions if you have any, and I could see, sorry, this is all unpublished. It's not the Polish presentation, but I would love to get any ideas that you may have any suggestions, yes. So in, oh, we're talking way, way along, this is in 1972, I did a study in Sweden, obviously, where this light day changes dramatic over the year. You can go from complete darkness to complete light. And it had to do with the need for the hospital system. And they need twice as many hospital beds in the middle of the winter for the safe populations even in the summer. So it was not only things you expect, depression, alcoholism, in the summertime, they would have half the incidence of myocardial infarctions, half the diagnosis of cancer. And at that time, it was a complete mystery. It has to be something in the area you're talking about. Have you got work that's looked at these greater health issues? So honestly, the most interest I get in my talk is people who either have lived in Northern Europe, like when I present to people from Norway or Sweden, they get so excited because they could directly. I have lived through two full Swedish winters, and I can tell you, that's brutal. I could tell you like you go to Finland in the summer and the amount of sunlight is just unbelievable. It's just incredible. You feel invigorated in the summer. It's amazing. I think there's a couple of reasons. I think a lot of people think it's all through the clock, and I have no doubt the clock plays a role. But I think there is a function for light that reminds me of our sleep need. We know we need to sleep, and we know without sleep we have a problem. I think there is something about photons hitting our retina and signaling to the brain that is just important on its own merit. And you think I'm crazy saying this, and I really feel crazy saying it, but I honestly think you cannot just simply explain it by the clock. Let me tell you why. Because as I showed you, you could entrain the clock at very dim light conditions. You could entrain the clock with a 50 minute light pulse at the right phase. But that doesn't alleviate the problem associated with depression, with lack of concentration, with alertness. So clearly there is a completely different system that we're starting to scratch the surface off. That why do you have IPRGC is projecting to 21 to 22 brain regions, including regions that connect to migraine. Region that connect, as I'll show you in my talk today, to the medial prefrontal cortex. So we know that ganglion cells connect to the visual cortex by the LGN. But there's this another cortex that IPRGC project to the center of your personality. IPRGC in two synapses reaches your medial prefrontal cortex. Why is that important? Why is activating this pathway important? How long we have to activate in a human to start getting the benefit? How many days? Because all these effects of light don't happen acutely. It's not like you go to a dark room and your rods get adapted. For people to start feeling the depression symptoms, they have to live for a long weeks under these dark. And when they go to travel to longer day, it takes them a couple of weeks to get over it. In fact, one of the saddest things I've heard about people who commit suicide is that they get so depressed in this winter that they can't even have the energy to do it. And when spring comes, they get the energy to do it because they are still depressed, but now they have the energy to. So actually, the suicide rate spikes up in the spring in Scandinavia. So a lot of these people are coming to me and they are building completely new hospitals now that are using the melanopsis system. And actually the Department of Energy just recently decided that in addition to vision that you have to include light health effect, they call it. I hope they stop calling it light circadian effect because I don't think that's true. I think it should be light health effect. And people think that heart rate cancer in nurses in Denmark have shown that nurses with a lot of shift work have higher incidence of breast cancer. So it's very interesting we're scratching the surface. Probably a lot of people think that's related to impact on the immune system. The immune system and the light and the circadian clock that affects the infectious diseases. Absolutely, the effects are no doubt incorporated. We tabulated that. You know, it's really interesting because every time we work on the nocturnal animal, people say, well, they are nocturnal, they hate light. And then you look at rodents in the wild and it's amazing. They have their burrows, right? But anytime they have a chance to go in the day and just be close to their burrows and just bask in the sun, they do that. I went to China and there's this rodent that just literally in the morning, they all like an army sit and look at the sun. And they are like, it's amazing. It's just people think rodents hate light. If you put mice in constant darkness, they form depression and anxiety-like behaviors. And you know, so there's this idea that nocturnal means hate light, but I think most of these animals, even like the blind mora, they still have ganglion cells that express melanopsin, they still have eyes under their skin and they still go out of their burrows every time to get some information about the light environment, even though they live underground. Yeah, I've got a picture. It was about probably minus 45 degrees and miserable. And all these people coming out of their offices just to catch the half an hour of the sun's on the, just to watch. That's amazing. It's just like you talked about as well. So they are changing. So what they are doing, they are doing light cycles where you get more blue, white and rich light and very high intensity in the day. And then you dim it to the red, which doesn't activate melanopsin at night. And in fact, they are saving 70% in energy because you don't have to keep the lights on. So people are changing and the Department of Energy are thinking of completely revamping schools, hospitals, offices, homes. You know, I follow this at my home and my sleep cycle have been great, but I'm also 100% sure I'm 100% placebo because I love this business. So I'm like, no matter what's gonna work for me. So thank you for this question. Thank you. So this is maybe a silly question, but do you know how your iPhone has that sleep cycle? Yeah. So that you can look at the screen and you can wake up worried about this person, you know? Right. So is that not blue or? So I could tell you, you could do the experiment. It's really, I tell my students, just put the iPhone at the dimmest light at night. You don't want to tell the difference and in the day you won't see the screen. So I think you should dim it to the, I mean, literally to zero. I wish they have even lower. I think if you dim it, the lowest part is much less effective, but like I see people using iPads at high light at night and you could see them walking in the street and this blue halo around them like, oh man, I have problems sleeping. And you're like, I wonder why. Your melatonin is literally zero and it's going to take two hours after you switch the light off to start going up. And you're literally completely alerted. You're possibly getting more light in your eyes than your office if you're in a cubicle and the light is more orange like this you see here, not the blue one, like this is more orange-ish. And you're possibly getting less light in your workplace and when you're going home shining the blue light in your eyes that your buddy thinking, oh this is the day that was the night so you sleep at work and you can't sleep at home. And you wonder why because your clock is shifted. You're literally a 24-hour shifted person without your knowledge. Because it's all subconscious, remember, you have no clue where you are, you just feel it, but you don't know what's happening. That's the problem. The guy that implemented the software and do a lot of computing splits is flying into town for summer's noon talk. So if you want to move with Michael, her. Okay. Very much, appreciate it, sir, thank you.