Scientist Stories: Li-Huei Tsai, Molecular, cellular, & circuit basis of neurodegenerative disorders

So it's my pleasure to introduce Professor Li-Huei Tsai from MIT today. It's a great privilege to do this, to connect kind of this whole seminar series in Professor Tsai's talk, because we began this spring talking about light as a scaffold for organizing whole biological systems. And so today, I think Dr. Tsai will give us a view towards looking specifically at whole brain function and how something that seems anomalous, if you're not familiar with her group's work, may be a real key to understanding some of these challenges that we've seen in dementia research, Alzheimer's disease research, and other clinical challenges that are related or relevant to quantum biology. So let's start with— in 1972, Philip Anderson, the Nobel Prize-winning physicist, wrote a nice little essay in Science called "More is Different." And in this, from which this figure on the left is taken, And so I kind of wanted to hearken back to Dr. Anderson's work because, you know, while Dr. Tsai is certainly more in the biological clinical arena, there's a strong connection between this idea that at every scale of physical reality as well as chemical and biological reality, we have a set of laws that in some sense can be derived from more fundamental laws, but in another sense don't need those fundamental laws. For example, to understand how to build bridges and build rocket ships, or even to put a helicopter on Mars and go up and down, we can create sort of models that coarse-grain or roughly estimate theories that we have from quantum field theory and quantum mechanics. We don't need to use those very refined, fine particulate understandings of reality in order to build these things for the most part. We can sort of coarse-grain into Young's moduli and shear and stress tensors, and then we can use those things to build things that are based on classical physics, right? And so Philip Anderson makes this point very strongly in the idea that um, just because we examine the fundamental theories of reality and we have to understand what the parameters of those physical theories are, we must understand that at every new level or scale, there's a whole new rich phenomenology that we must understand and appreciate. And I think Professor Tsai's work really fits in there in the understanding of the brain. On the right, this is from a book chapter we wrote 2 years ago, looking at kind of how the world of quantum fluctuations at the scale of the attosecond, the picosecond, the femtosecond— how could these quantum fluctuations be implicated on the scale of the brain, at which the scale that Professor Tsai is going to talk about today is really at the hertz scale. So this is, you know, 18 orders of magnitude that are stretching from these quantum fluctuations to the brain, going through biomolecular complexes to neurons, neural networks in the brain. And so, like Stephen Hawking, who popularized this theory of everything, but then on his deathbed sort of recanted the hubris of such a theory, understanding these emergent phenomena will become critical for understanding that, yes, you can describe fundamentally quantum phenomena in classical manners. And understanding what is truly quantum that cannot be described classically from what is able to be described in a coarse-grained classical manner is crucial. But it's also important to recognize that these phenomenologies that are happening in the quantum world that are, let's say, non-unitarily or reducing themselves into classical phenomena that are measurable by biologists and clinicians, that we recognize that there is an intimate dynamism between the two worlds. So in 1932, Niels Bohr gave a lecture on light and life. And so we're going to take the idea of how do light stimuli, or how do external photonic stimuli affect living processes, specifically in the brain? And so by now, I know many of you have seen this, this picture, which ultimately looks kind of at a bottom-up approach of the neuron and the brain. And so we have this kind of working picture of this intimate nexus between the mitochondrial reticula, and the cytoskeleton, composed of not only actin filaments and intermediate filaments, but also microtubules. And in this intimate dynamism, we think of these biophotons that are potentially being emitted from free radicals that are emitted from mitochondria. But what Professor Tsai's work, I think, makes us consider is what are the driving processes from a top-down perspective? Because if— and I'll present a little bit about what her work reveals to us, but this idea of an external non-invasive flicker that is visually rendered but then can stimulate neuronal networks to then change morphology of those networks could then possibly stimulate networks of different aromatic molecules that could then be implicated at the quantum level. So there is possibly, from a bottom-up approach, things we can do to modulate, enhance, alter. But today we're really going to focus on the alternative view, which is from the level of the brain, the visual stimulus, and the neuron, how might these things be implicated at the biomolecular level? This is a snapshot from the Human Brain Project. As you can see, and as I've mentioned, we have in the brain these very divergent spatial, temporal, and energy scales. Today, Dr. Tsai is going to be right in the middle of this dynamic hierarchy or milieu. And she's going to really focus on this center regime of the Hertz scale, but then also to understand some of the biomolecular downstream effects that are a result of this amazing finding. And so now that I've alluded to it, in 2018, this was a Nature paper that Dr. Tsai's group, and in collaboration with Edward Boyden, Emery Brown, and some others, have found that this particularly 40 Hertz frequency this flicker in Alzheimer's disease mice was able to reduce amyloid load through this so-called gamma entrainment. And so there were a number of downstream effects. And I hope today Dr. Tsai will tell us a little bit about some of the clinical trials that have emerged at the human level from these experiments in mice. But these downstream effects include an increase in spatial learning and memory for the mice. A loss or a decrease in synaptic loss, and an increase in vesicle trafficking in the microglia and the neurons. So this is really exciting from the level of a physical stimulus that's non-invasive that can entrain the brain. And it raises questions about even things like meditation and other types of phenomena that have been historically associated with such gamma synchronies. So just to give a sense of how this complicates our view of artificial intelligence and consciousness. So I've presented this in my introductory talk, but this is kind of a lower computational bound for the human brain's computational capacity, given if you look at the neuron as the basic computational unit. If you go into the cytoskeleton and you assume only the, you reduce the entire cell to a single type of protein, you increase that computational capacity by 10 orders of magnitude. And if you use atoms, and you're even more exhausted, you increase that by another 10 orders of magnitude. And so we're looking really at advancing not just the nature of consciousness as saying, OK, we can take this consciousness, download it into an intelligence, but it also complicates our view of the complexity of the interactions in this light and life feedback loop. So something to think about are the energetic considerations of what's happening in Dr. Tsai's work, but then also the classical versus quantum mechanisms that may be at play in the brain that if it were scaling according to our classical computers, we would have several nuclear power plants worth of energy output between our ears. So let's move from these partitions that we have in sort of a reductionist science to this kind of wholeness and interconnectivity that Hermann Weyl and other founders of quantum mechanics and other scientists like Dr. Tsai have presented us. So Today, I would just like to say that Dr. Tsai is the Picower Professor of Neuroscience at MIT, the director of the Picower Institute for Learning and Memory. Her lab investigates disorders of memory and cognition, in particular neurodegenerative conditions such as Alzheimer's disease, at scales ranging from the molecular and cellular to that of circuits and systems. So her work has highlighted important roles in neurodegeneration for the dysregulation of chromatin and genomic integrity, as well as the kinase CDK5. So this study investigating dysfunction in neural circuits has led to the discovery that sensory stimulation of 40 Hz gamma oscillations and the associated synchrony produces widespread ameliorative effects, including a reduction in amyloid and tau pathology. So she has demonstrated that the enhanced gamma power recruits the brain's immune cells and the vasculature. Dr. Tsai is a fellow of the American Association for the Advancement of Science, the National Academy of Inventors. She's a member of the National Academy of Medicine and an academician of the Academia Sinica in Taiwan. She's a recipient of the Mika Salpeter Lifetime Achievement Award and the 2018 Hans Wigsell Research Foundation Science Prize for her research on Alzheimer's disease. We are very privileged to have her with us today And I ask her to take it away. Thank you so much, Dr. Tsai.

Dr. Tsai, you're muted at the moment.

Thank you, Philip. That was quite an introduction. So special. I really appreciate it. So it's absolutely a pleasure to be here to present our work. So if I can share my screen. Great. Can everyone see my screen?

Yes.

Yes, that's good.

Great. So as Phillip told you earlier, today I'm very excited to tell you about a completely non-invasive kind of brain stimulation protocol that we developed in the lab. We found that this non-invasive stimulation basically using light, and later on we also integrate sound into the stimulation, really is beneficial to the brain in animal models of neurodegeneration. But recently we started to have some very preliminary data from human subjects that I would like to share with you. The human data is not published. So this is my disclosure slide. I would like to particularly point out that Cognito Therapeutics, a small biotech company, licensed our technology technology and is currently conducting larger-scale human trials.

So, um, could I just interrupt you? There's something, um, rubbing up against your microphone.

Oh, okay, okay. I will try to fix that. Okay, well, is this better now?

That's far, far better. I just didn't want the whole lecture to go through, but thank you very much. Sorry to interrupt you.

Great. So I just want to have a few words about dementia. So with the extension of lifespan, obviously age-related diseases have risen rapidly, and the increase in the incidence of dementia is particularly alarming. Today, according to a 2015 report, 47 million people are living with dementia, and this number will double or triple in every continent in the next 3 decades. And the majority of the dementia cases are caused by Alzheimer's disease. Alzheimer's disease is considered the most pervasive and urgent problem facing our modern society, and we do not currently have effective treatment. So we know that Alzheimer's disease is complex and takes at least 20 years to develop. It has very defined pathological features in the brain. As many of you have heard, there are abundant amyloid plaques or senile plaques, as well as neurofibrillary tau tangles. But there are also other features such as loss of brain volume, shrinkage of the brain, loss of synapses. And initially it precipitates in forgetfulness and other cognitive impairment, but eventually leads to dementia. And over the past several decades, research has revealed many, many cellular and subcellular defects that happen during the disease. And recently, people also realized that the disease is not just about the nerve cells known as neurons, but many other cells in the brain, other types of cells in the brain are also affected. Then there is this genetic component of the disease as well. In addition to a few of the genes with variants that can cause Alzheimer's disease, that run in the family, there are also 2-3 dozens of other genetic loci in the human genome that can significantly affect one's risk of developing Alzheimer's disease in their lifetime. But despite all of this knowledge, we still don't completely understand the disease. And as I said, there is no effective treatment. And in fact, in the last 15 years or 20 years, none of the new interventions succeeded in human trials. So among other things, my laboratory has been studying neural circuit regulation and network activity in Alzheimer's. Today, I would like to focus on a particular neural network activity known as oscillations or rhythms. I propose that an impairment of the 40-hertz gamma rhythms and its co-activation across various brain regions undermine the brain's response. So let me just tell you a little bit about brain rhythms. In the mammalian cortex, there are at least 10 different classes of oscillations that cover more than 4 orders of magnitude in frequencies, from the infraslow to the ultrafast. So, as you heard from Philip, this Hertz, meaning how many times the neurons can fire per second. Okay. And one thing about all of these different frequencies of brain rhythms is that it's highly evolutionarily conserved. Throughout evolution. So if you look at all these different frequencies of oscillation, you can detect in very small mammals such as mice and rats, and even in the human brains, which is clearly much, much larger, but the frequencies still remain. So I think these facts really suggests that all of these different frequencies of oscillations, they all play important role in brain function. And as Philip pointed out and I alluded to earlier, we are particularly interested in the so-called gamma oscillations, which ranges between 30 to 80 Hz. So just a very, very brief history of neural gamma oscillations. So neuronal oscillations actually were first observed more than a century ago as a fundamental brain activity in rabbits and in monkeys. And in the '20s and '30s, Hans Berger and others classified brain oscillations, including those in humans, by their frequency, which range from infraslow to ultrafast. And gamma oscillations were first reported in 1942. And identified in many brain regions, first in the olfactory bulb of hedgehogs. And the scientist reported this is named Agar Adrian. And further progress in the field led to association of rhythms with many different behaviors and an explosion of research in the last 30 years, including seminal work from a German neuroscientist, Wolf Singer, has produced a theory that gamma oscillations synchronize within and across connected brain regions. And he and Yuri Bussaki further proposed that gamma temporarily organized spike timing with functional consequences. Today, we know that gamma rhythms are highly specific to behavioral state and key to information processing. Enhanced gamma power in the brain is observed when human subjects or animals are engaged in higher-order functions such as working memory, sensory processing, and spatial navigation. Meanwhile, we still don't completely understand gamma. For instance, we don't know what are the cellular and molecular substrates of increased coherence associated with gamma. We don't know how neurons respond to gamma. We don't know whether gamma impacts other brain cell types. So basically, we don't know what are the cellular signatures of gamma in the brain. One thing that has been reported over the years is that reduced gamma power is detected in multiple mouse models of Alzheimer's and also in Alzheimer's human subjects. I'm showing you an example of a piece of data from my laboratory where we recorded brain rhythms in healthy wild-type mice or in an amyloid Alzheimer's disease mouse model known as the FADx FAD model. And we specifically look into very young animals, 3 months old. So in this Alzheimer's model, 3 months old is presymptomatic. They don't show any behavioral deficits or even overt pathological manifestation. So, when we recorded the rhythms in the brain, we found that, for instance, in the healthy animals, we see a very strong power of gamma oscillations, as you can see here, the frequency range in the gamma range. But when we look at the 3-month-old, amyloid model, we found that there is already a very significant reduction in the gamma power of oscillations. So this was very surprising because this, as I said, this is the presymptomatic stage of the animals. So this observation also motivated us to really ask, since the reduced power of gamma happened this early in the course of the disease, does it contribute to the later manifestation of the various pathology and symptoms? And these are a partial list of the literature showing reduced gamma power in the many different Alzheimer's models, including amyloid model, tau transgenic model, as well as a sporadic Alzheimer's disease risk gene, ApoE4 knock-in animals. So knowing that gamma power is reduced in Alzheimer's disease, people have attempted to increase gamma power in the brain. And as you mentioned, that a publication that really caught our eyes from Jorge Palop's group at UCSF. They also look at another Alzheimer's disease amyloid transgenic mouse model. They call it HAppJ20 animals. In these animals, they also show Similarly, the power of gamma oscillatory activity, especially centering around 40 Hz, is very much reduced. So what they did was to genetically manipulate a particular type of neurons known as inhibitory alpha-mine-positive neurons by overexpressing a voltage-gated sodium channel into the cells. They showed that this manipulation greatly enhanced gamma oscillation power. This is associated with improved cognitive function in this model. We initially decided to see whether in our hand, if we increase gamma power is also beneficial to the Alzheimer's disease mouse model. So we use a different technique known as optogenetics, which is very, very invasive. What this involves is to, first of all, use a viral vector to precisely introduce an opsin into a particular part of the brain known as the hippocampus. Hippocampus is the part of the brain known to regulate learning and memory function. We then stick a fiber into that brain region so that we can use laser light to activate the opsins in the neurons in the hippocampus. This way we use 40-hertz laser pulses to introduce into the hippocampus of these mice. What we can see is that when we perform electrical recording, we can see that the 40 Hz gamma power is very significantly increased after this manipulation. Okay, so this is not surprising because this kind of manipulation has been reported before. But what was really surprising and in fact came as a shock to us was the following observations made by my former graduate student, Hannah Alcarino. So what she found is that in this amyloid model, in the hippocampus, even though in this young age we don't see overt amyloid plaque pathology, but the components of the amyloid plaque known as the beta amyloid peptides— those are the small peptides, they can aggregate, they can form fibrose, and they can precipitate into the plaques. When we look at the components in the hippocampus of these mice, we found that something is going on, because we found that those mice receive 1 hour of the 40-hertz stimulation show about 50% reduction of the abundance of this beta amyloid peptides. There are two different versions, A-beta 1-40 and A-beta 1-42. We found that with 40-hertz stimulation, both peptides were cut roughly by half. Moreover, if we use a different frequency such as 8 hertz or use a random frequency stimulation, None of these manipulations reduced beta amyloid peptides. So only 40 Hz frequency reduced beta amyloid peptides. So I think you have to appreciate the surprise that came with this observation, because up until our experiments, everyone else All the literature you can find, when they talk about brain rhythms, different classes of brain rhythms, they describe it in extremely computational terms, okay, or systems neuroscience terms. Brain rhythms, information processing speed, and maybe some sort of association with behavior states. But nobody, nobody has ever shown that these brain rhythms can actually change the biochemical aspects of the brain cells. Okay, so, so because of this, in fact, many people in the field of neuroscience are extremely skeptical when it comes to brain rhythms. They claim that brain rhythms are just epiphenomenon. They don't actually do anything. They are not important. But our observations for the first time show that's not true. Brain rhythms matter, and at a particular frequency, it changes the biochemical processes of the cells. So when we presented these results in our department, of course my colleagues were very, very surprised. And some are more optimistic than others. So I remember Emery Brown, a brilliant computational neuroscientist, and he himself is also a physician. He came to me and said, Li Wei, look, your observations may have tremendous translational potential, but this optogenetic approach that you have to introduce a virus into the brain and stick a fiber optic, it's just too invasive. It's never going to be done in humans. You've got to think of a way to manipulate the brain rhythms non-invasively. So he challenged me. So, you know, we, my graduate student Hannah Alcarino and I, went back and then, you know, thought about what can we do. So the only thing we can do is to go back to the literature and to see whether anyone has ever tried to manipulate brain rhythms using non-invasive way. And really, again, Wolf Singer came to the rescue. So in this 1998 publication, Rager and Singer, they performed the following experiment. So they actually shine light to cats, and they manipulate the light to flicker at different frequencies. And simultaneously, they also perform electrophysiological recordings in the visual cortex of the cats. So the visual cortex is part of the brain important for processing visual input to downstream brain regions. So this is the results they show. So they presented the light flicker anywhere from 0 to 50 Hz, and then they recorded the response in the visual cortex. And what you can see is absolutely a linear relationship. Meaning if you stimulate at 10 Hz, then you can record 10 Hz oscillation in the visual cortex. 20 Hz, 20 Hz. 40 Hz, 40 Hz. So we sort of all of a sudden, you know, we realized that at least in cats, it is possible to manipulate oscillations in the brain just by giving them different visual inputs. So the next question is whether this is going to work in mice, because we know cats have great vision, but mice, they are not well known for their good vision, right? So, so Hannah did the following thing. So she house the animal, the mice, in a cage, and then she surrounded the cage with LED light strip. And then she used a very simple Arduino system to manipulate the light to flicker at different frequencies. So this is the experiment. And then during the stimulation, we recorded the neural activity in the visual cortex of these mice. And then there you can see when we presented 40 Hz light flicker, then what we saw was a beautiful increase in the 40 Hz oscillation power in the visual cortex. So, in fact, this sensory stimulation-induced brain rhythms seems to be pretty universal. It works in cats, and it also can work in mice. So we turned this approach, gamma entrainment using sensory stimuli, or genus. So, the next thing, of course, we wanted to know whether the sensory stimulation-induced brain rhythms can also lead to beta amyloid peptides reduction. So, Hannah performed the following experiment. Again, she exposed the mice just to the live flicker for an hour, with different frequency of this light flickers. And then she harvested the visual cortices, which is in the back of the brain. Okay, this is the olfactory bulb, this is the back of the cortex, and this is the hindbrain. And then we measure the abundance of beta-amyloid peptide 1-40 and 1-42. And from the figures, I think you can appreciate Again, we saw with 40 Hz light flicker stimulation, we saw a huge reduction of these amyloid peptides. And none of these other frequencies we used, including constant light, 20 Hz, 80 Hz, or random light frequency, none of them significantly reduced beta amyloid peptide abundance. So the next question then is the following. So, so far I showed you that using sensory stimulation, namely light flicker, can entrain neurons in the visual cortex to produce rhythms. Okay, but we know that a lot of the brain disorders, such as Alzheimer's, affect many different brain regions, including the hippocampus. And in fact, the sensory cortices are not known to be particularly affected in neurodegeneration. So the question is, whether this light flicker-induced oscillations are very local or it can be propagated to other parts of the brain. So to address this question, a postdoctoral fellow, Chinnadhaikian, who is a fabulous electrophysiologist. He performed the following experiment. He produced a custom-made microdrive that can simultaneously target different regions of the brain. So he sticks these electrodes into an awake, freely moving mouse. He then turned on the light flicker to the mouse and then recorded the neural activity in the different parts of the brain. Okay, so this is a— so you can see the heat map of the neural activity in the spectrogram. So this is LED light occluded and then LED light exposure. So as soon as we turn on this LED light and expose the mice, you can see there's a sharp increase of 40 Hz oscillation power in the visual cortex, which is not surprising. I told you earlier already. But you also see increased 40-hertz gamma power in the somatosensory cortex, in the hippocampus, which, as I told you earlier, controls learning and memory, and in the prefrontal cortex, which is the part of the brain in charge of higher-order brain functions such as attention, executive function, decision-making, and so on. So this experiment would suggest that, in fact, light flicker-induced neural oscillation can propagate to many different parts of the brain. Moreover, another finding that China made, which I think is very interesting, is that he looked at co-activation of two different brain regions, such as co-activation between the visual cortex and the hippocampus, or the visual cortex and the somatosensory cortex, and so on, at 40 Hz. And what he found is very significant increase of co-activation among many of these different brain regions. So we call it coherence. And the increased coherence usually translates to increased speed of information processing. So information in the brain can be processed with increased speed. Then subsequently, Tina asked, whether— so, so far I've shown you all of the exposure, which is always acute, such as 1-hour exposure. So, Tina wanted to know whether further beneficial effects can be achieved if he chronically exposed this Alzheimer's model mice to genus. So this is the following experiment. He now used a much older amyloid model, 5CFAD mice, 9 months old. And by this age, you can see a very extensive precipitation of this famous Alzheimer's pathological feature, amyloid plaques. So you can see this magenta punctate kind of signals throughout the mouse brain in the cortex and in the hippocampus. Then he exposes mice to 40-hertz live flicker 1 hour per day for a total of 21 days. He then measured the density of this magenta amyloid plaques. And what he found is significant reduction of the abundance of these plaques, not just in the visual cortex but also in the somatosensory cortex, in the hippocampus, and in the prefrontal cortex. So it's not just that the gamma power can propagate to the whole brain, but it's beneficial effect in reducing beta amyloid plaques can also propagate to many parts of the brain. So far, I exclusively talk about the amyloid. And you may wonder, there are many other Alzheimer's pathological features such as neurofibrillary tau tangles. Is genus also beneficial to remove that? So here, Gina worked with the tau transgenic model known as the P301S/P19 mice. So these animals overexpress or produce a huge number of the tau protein. It's a variant version of the tau protein identified in patients with frontotemporal dementia. So you can see in this quite— this older tau mice with no stimulation, you can see this, this green signal very abundant in these nerve cells or neurons. So this is accumulation of this pathogenic tau protein. And you don't really see this in a healthy wild-type control animal. Remarkably, with 1 hour per day, 21 days of exposure to the life liquor, I think you can appreciate there is a profound reduction of this pathogenic form of the tau protein. And Tina again quantified in different brain regions, and he was able to show this beautiful reduction of the tau protein in all of the different brain regions that he evaluated. So the light flicker exposure at 40 Hz not only removes amyloid, it also removes tau pathology. A third feature of Alzheimer's I alluded to earlier is loss of brain volume. So Tina now turned to a a third transgenic mouse model that is known to show tremendous loss of brain volume and loss of the density of the nerve cells. So this is known as the CKP25 mice. So without stimulation, you can see this animal show great expansion of the ventricles. In the brain, which is also a very strong feature in human patients with Alzheimer's disease, and the thinning of the brain structure compared to healthy control animals. And then again, when he exposed these mice for 1 hour per day and a total of 6 weeks, You can see the expansion of the ventricle is ameliorated. And the brain structure, you can see a recovery of the thickness of the cortex. And further, if you look at individual neurons here, as shown as magenta here, you can see the density of this neurons is reduced without stimulation but restored after stimulation. So we can see multiple beneficial effects of light flicker stimulation in many different neurodegeneration mouse models. So, so far I show you that it is possible to use light flicker or visual stimulation to entrain neurons in the brain. But we also know that we are endowed with many different senses. So one of the questions we would like to address is whether we can take advantage of the other senses to entrain neurons. And I'm going to show you a couple of pieces of results that will show that it is also possible to increase the power of oscillations using sound. So this is a very close collaboration with Ed Boyden's laboratory, and especially with his former graduate student Hojun Suk, who really helped develop this algorithm for auditory stimulation, especially for mice, because mice have a very different hearing range from humans. So we came up with this algorithm using 10 MHz tone presented at 25 millisecond intervals. So this is 40 Hz per second. And we found that when we present this kind of sound stimulation to mice, then in the auditory cortex, in the hippocampus, we can rarely detect this regular 40 Hz firing of neurons in these brain regions. But more importantly also is that in the amyloid mouse model of Alzheimer's disease, when we look at amyloid plaques as shown in green here, we found that with 40 Hz auditory stimulation, there is a profound reduction of this amyloid plaques in both the auditory cortex and in the hippocampus after chronic exposure to this auditory stimulation. What about we combine light and sound, whether we can see additional beneficial effects? The short answer is yes. I'm not going to show you all of this data that we documented with this combined light and sound simulation, but I'm going to just use these movies to illustrate my point. What the combined stimulation does, which is better than a single modality stimulation, and later on, I think you can also see it in our experiments in humans is that it really helped to propagate the 40-hertz gamma power to different brain regions in a more timely manner. So we can shorten the time it takes to achieve the reduction of amyloid pathology. So here we collaborated with Kwanghun Chung, where he developed a method to clear the entire brain. So we rendered the brain transparent, and then we came in to use amyloid antibody to identify all of the amyloid plaques. So here, with no stimulation, you can just you know, all of this punctate kind of amyloid plaques distribution throughout the entire brain, in the cortex, but also in the deeper parts of the brain. Okay, but with auditory plus visual stimulation, what we found is the whole-cell reduction of these amyloid plaques, not just in the cortex and in the hippocampus, hippocampus, but also in the deep part of the brain. So we can see a whole brain reduction of amyloid plaques, very significant reduction with just 1 week of 1 hour per day of auditory plus visual stimulation. So now I show you all of this reduction of pathology. But what about function of the brain? And I'm going to just show you one piece of data. And we have done the experiments testing cognitive function using many different behavioral tasks. And in general, I can tell you that multiple strains of the mice it doesn't matter whether it's amyloid or tau or degeneration or even aged mice, after they are exposed to light or sound or combined stimulation chronically, they show very nice reduction of the level of anxiety., and they appear to be much more calm. And when it comes to learning a memory task, I'm going to tell you especially about these results using the so-called Morris Water Maze behavioral task, which is to evaluate the spatial learning and memory of the mice. So what you do is to put a mouse in a swimming pool filled with murky water. And mice inherently, they don't like to be in the water, so they will quickly look for escape. So, so what you do is to put a platform submerged underneath the surface of the water so they cannot see where the platform is. But if they swim Swim around, they will eventually find the platform. So we placed the mice to the swimming pool on a daily basis and initially looked at the time it took for the mouse to reach the escape platform. Okay. And I think you can appreciate in all of the different mouse models, when we place these mice in the swimming pool. So this is the curve representing with every day of training the time it takes for the different mice to reach the platform. And I think the important thing to see is with the daily training, the mice should take less and less time to find the platform if they learn. To locate the platform. But with 40 Hz stimulation group— so we have 3 groups: no stimulation, 40 Hz stimulation, or random frequency stimulation in this particular experiment— only the 40 Hz group, they show very significant reduction in locating the platform. And this difference is significant when compared to the no-stim group or the random simulation group. Furthermore, on the last day, what we did was to remove this platform altogether and place the mouse back to the swimming pool. So the pool mice, they're going to, you know, if they remember in this particular quadrant of the swimming pool the platform used to be there, they tend to spend a lot more time here to look for the platform. We basically quantify the time the mouse spent in the different quadrant of the swimming pool. You can see the 40 Hz group spent significantly more time in this quadrant than these other groups, indication that they already remember where the platform used to be. In all of the different mouse models, we consistently see the 40 Hz stimulated group performed better, showed better spatial learning and memory. Now, so I show you a whole bunch of beneficial effects. But my laboratory is really a lot more a mechanism laboratory. We want to understand how does this all happen. I can just tell you now, I don't have all the answers. I think we are getting there. What I would like to do is to share with you what I do know. The brain harbors many different cell types. The nerve cells or neurons is just one cell type you see in the brain. The brain also has microglia, astrocytes, oligodendrocytes, and blood vessel cells. And these microglia in particular are an extremely interesting group of cells. They are actually immune cells. They are the brain's innate immune cells. They constantly send out these processes to survey the environment. And if they sense that there's some injury happen to the nerve cells, they will come in quickly to remove the debris. And if they can sense any metabolite or other waste in the brain, they will also come in to remove that. However, under chronic toxic environment, this microglia would transform into an inflammatory state, and they will become a factory, keep releasing toxic cytokines and chemokines, and do further damage to the brain. Okay, so here I'm going to show you some short videos of how this microglia look. In a healthy mouse brain or in the CKP25 mice. Remember, these mice have a lot of neurodegeneration, loss of neurons, and brain atrophy. And then the CKP25 mice after 40 Hz chronic stimulation. So now you can see this microglia. They show this nice ramified morphology. And they show this tiling kind of distribution. So they don't really invade into each other's boundary. Okay, now is the microglia, how they look in the neurodegeneration state without 40 hertz. Now you see they totally lost this beautiful ramified or tiling distribution. And here, what you see is a cluster of microglia. They pile right on top of each other and look totally not like microglia anymore. Then finally, here is what the microglia are like after chronic 40-hertz live flicker stimulation. They pretty much revert back to this normal ramified-looking piling distribution. So you can see this, you know, again, you know, I think a lot of people told me that when they see this kind of data, it's a watershed, watershed moment, because this is the first time anybody ever showed that this kind of neuronal oscillation phenomenon can actually change the morphology and very likely function of different cell types in the brain. So we wanted to know further how the microglia respond to this chronic stimulus stimulation. So this is a very molecular experiment. I'm going to just at a very high level tell you the idea of this experiment. So what Tina did was after quanine stimulation, he was able to purify this microglia from the brain. And then he performed an experiment known as RNA sequencing. Basically, it's to look to see whether there are changes in gene expression in this microglia. Because if there are changes in gene expression, it's most likely going to change the function of the microglia. So, we have several groups of animals. So, we have the CKP25 mice without stimulation and compared to a healthy animal, just to see how the microglia change in neurodegeneration. So, what you see is an increased expression of genes involved in immune response, inflammation, and apoptotic pathway, which is cell death, and downregulation of genes involving motility and other homeostatic function of microglia. But now, after 40-hertz stimulation, we compare the treated animals versus untreated animals. And here, what we see is reduce expression of this immune response, inflammation response genes, and then increase expression of membrane trafficking, vesicular transport, and catabolic process genes. So basically restore motility and homeostatic function of microglia. So basically this experiment this gene expression signature experiment is very consistent with the morphological analysis of the experiment that I showed you in the previous slide, which is this stimulation can restore normal microglia function. And I'm going to skip this slide. So, just to come here also. So, Tina also purified nerve cells, so neurons, from treated animals and looked at gene expression changes. And what he was able to show is that the stimulated animals show tremendous increase in the expression of chemical synaptic transmission and dendrite development, long-term potentiation, vesicular transport, synaptic transport category of genes. So basically, increased synaptic function as well as plasticity of the synapses. In addition, he also used immunostaining to look at literally the density of synapses and showed that while in the neurodegeneration model the synaptic density is reduced,. But following chronic treatment, synaptic density is restored. So this 40-hertz exposure is very beneficial for synaptic function. And microglia and neurons are not the only cell types responded to GINAS. Anthony Martorell, my former graduate student, found that astrocytes, which are known to be the type of glial cells that normally support the function of neurons, their number can be increased following 40-hertz stimulation as well. So astrocytes also respond to the increased 40-hertz oscillation. And finally, not just microglia, astrocytes, and neurons. Turned out that the blood vessels in the brain also responded. This is just phenomenal. The brain is full of vasculature, especially capillaries. They are extremely abundant in the brain because our brain cells need a lot of energy in order to support their function. This abundant capillaries and vasculature constantly supply oxygen, and glucose to support the energy needs of the brain cells. And so what Anthony found is that following 40-hertz stimulation, the overall diameter of those capillaries is very significantly increased. Okay. And later on,