 Good morning, everyone. Welcome to the first Purdue Engineering Distinguished Lecture of Academic Year 2023-2024. My name is Nankong. I'm the Professor Interim Head of Biomec Engineering. So this very first lecture of the academic year is hosted by the Welland School of Biomec Engineering. It's our distinct honor to have Dr. Huda Zakbe to deliver this lecture. I've spoken with Dr. Zakbe and I told her we have the finest faculty and students and many of them are here. So welcome. A little bit information about the Purdue Engineering Distinguished Lecture. Beginning in 2018, this lecture series has been inviting world renowned faculty and professionals to Purdue Engineering to encourage thought provoking conversations and ideas with faculty and students regarding the grand challenges and opportunities in their fields. So we're here and we're eager to hear the grand challenges and opportunities Dr. Huda Zakbe is going to present. Now I'm going to turn to Dr. Tamara Kinzer-Ursim, Associate Dean for Graduate and Professional Education for the College of Engineering. And she's also one of us on our faculty. And Tammy, please come up to introduce Dr. Zakbe. Thank you, Dr. Kong. It is my very great pleasure on behalf of the College of Engineering to welcome Dr. Zakbe here today. Dr. Huda Zakbe is a professor in molecular and human genetics at Baylor College of Medicine with additional appointments in the departments of neuroscience, pediatric section of neurology, translational biology and molecular medicine and developmental neuroscience. She is the Ralph D. Fine MD endowed chair at Baylor and the founding director of the Jan and Dan Duncan Neurological Research Institute at the Texans Children's Hospital and a Howard Hughes Medical Institute investigator. I really enjoyed meeting with Dr. Zakbe last night at dinner and I think you'll agree with me that her path to these current positions is incredibly inspiring. She originally wanted to study literature but was convinced by her mother to study biology as a more to have a more stable and independent career growing up as a woman in Lebanon. She ultimately obtained her bachelor's of science degree from American University in Beirut and attended medical college there for one year before civil war broke out in Lebanon disrupting and destroying so many lives. She ultimately made her way to the United States with family and attended a medical degree from Mary Medical College which is an HBCU near Nashville Tennessee. She then joined Baylor College of Medicine for her residency and subsequently postdoctoral training in molecular genetics and has been at Baylor ever since. Dr. Zakbe's expertise ranges from neurodevelopment at the beginning of life to neurodegeneration which we all know just happens towards the end of life. She and Dr. Harry Ott discovered that spino cerebral ataxia type one is the gene that is caused caused an expansion of a polyglutamate sequence inside the protein that causes aggregation of that protein inside neurons. This has had profound consequences for other diseases in which protein aggregation in neurons causes their pathogenic function and has implications for studying Parkinson's disease, Alzheimer's disease and other neurological disorders. Her work in neurodevelopment has led to the discovery of the Math 1-821 gene where she showed that it's involved where she showed its involvement in prosception, balance, hearing, vestibular and respiratory function. She has also discovered that mutations in the MCP2 gene cause the neurological disorder RET syndrome which is the story that we're going to hear about today. Now I mentioned that Dr. Zakbe is the founding director of the Jan and Dan Duncan Neurological Research Institute and she's really been the lifeblood of this institute building it from the ground up with the mission to foster multidisciplinary collaborative environment to facilitate fundamental discoveries in neuroscience with the ultimate goal of translating those discoveries into therapies for the many incurable neurological and neurodegenerative conditions. The institute currently has more than 30 principal investigators and working collaboratively they have discovered more than 40 disease-causing genetic mutations and have launched promising clinical trials in RET syndrome and tuberous gloriosis complex. Dr. Zakbe has trained over 100 scientists and physician scientists including two biomedical engineers and two professors here at Purdue. She has been elected to the National Academy of Medicine, the National Academies of Science, the American Academy of Arts and Sciences. Dr. Zakbe's honors include among others that I don't have time here to list the Shaw Prize in Life Science and Medicine, the Breakthrough Prize in Life Science, the Canada Gardier International Prize, the Lundbeck Foundation 2020 Brain Prize, and the Cavley Prize from the Cavley Foundation. The College of Engineering is very excited to have such a distinguished researcher and thought leader here with us today so please join me in welcoming Dr. Hudak Sogobi. Thank you so much Dr. Kuzer Ersem for this really kind and detailed introduction, now they know everything about me more or less. It's really a pleasure to be here. There's always something special when it's your first experience and this is my first experience speaking at the School of Biomedical Engineering. So you're going to be my judge and you're going to teach me something. You're going to tell me if I did a good job or bad job. How can I do better? I've truly, truly had the joy mentoring two biomedical engineer. One came from Georgia Tech and then UCSF and the other came from UVA and I got to appreciate how important it is to really have engineer in the midst of biologists and to have people trained in both disciplines or at least collaborate. The impact on science will be great so I look forward to see what you will accomplish. So as you've heard from Tammy today I'm going to tell you about my work on red syndrome and the reason I chose to talk about this work is really multiple things. It's a pretty good disorders to learn as much as you want to learn about childhood neurodevelopmental disorders. It's a common problem about one in 40 children born today have autism or intellectual disability. So more we understand about the brain and how these childhood diseases happen the better it is. Second, it's sort of one of those disorders that we're doing work that I think has some fertile grounds for your imagination to see what you can do for this disease. So that's the reason I chose this disorder. So red syndrome is the reason I'm a scientist. I was trained as a clinician, pediatricologist. I was in my first year of pediatricology training and thought I'm going to be a clinical pediatricologist until I met Ashley. And Ashley came to us with this history. She was perfectly normal healthy child. You see her here sitting straight holding a book with her hands and leaving through the book so she can use her hands. She's doing what an infant should do. You see her here riding a rocking horse. Notice how she's using her hand to hold on the horse. That's very appropriate for her age. She continued to do well for about two years and right around two years she stopped running to the door to say hello to her dad when he came from work. Her dad is a professor at Texas A&M University and he got concerned and then all things changed. She stopped using her hands. She couldn't hold anything with her hand. She stopped walking normally. She was shaking. She lost language. She lost social skills. Everything she knew how to do. She used to sing nursery rhymes. That disappeared. And when I met her, as you'll see in this picture, she's constantly wringing her hands. She can't use her hands anymore. So she clearly had a period of normal function and development and then deteriorated. And that's really what interesting me, red syndrome. I used to think as a pediatric neurology in training that there are two types of diseases. Developmental. You're born and the wiring of the brain is abnormal. A part of the brain is missing. Things manifest at birth. Or degenerative. You're born normal and things deteriorate. But red was neither nor. The brain looks normal at birth but they lose all these milestones and there's no degeneration. And the month I saw Ashley, which would be 40 years this October, it was 1983 in October, that same month a paper came from Bank Hagbert in the annals of neurology describing for the first time in the English language a syndrome called red syndrome. And she matched all the features of the syndrome. At that point, only European patients had been seen. He reported 35 cases from Europe. Andrea's red was the first to see these girls and describe it. But it was in German in 1966. None of us read the paper. But when it came in English, we read the paper and we noticed. And science always remembers serendipity is important. But if you're prepared for an observation, that's where magic happens. It just happened that the same month I saw Ashley a week later, I saw a nine-year-old girl walking into the clinic with a diagnosis of cerebral palsy ringing her hand. Having seen one girl with red, I immediately recognized the second one. So that's really what drew me in. Seeing two in one week from a syndrome no one heard about, I realized this is really important and I want to study it. And in this video, I just want to show you red syndrome in action. And you can see here the hand-ringing movement. That's really a hallmark of this disease. And you'll see rocking activity, features of autism. And then you're going to see that when she stands, she's freezing. She has all the capabilities, all the motor strength to move. But she can't plan that movement. She just and when she finally does it, she's unsteady and she falls back. So having seen two in one week, I asked my clinic volunteers to find me charts of any girl with some features of red and they found me six charts. And so within a month, I became the country's expert on red syndrome. So the patient wrote a paper in the New England Journal of Medicine. And when this was published in 1985, patients started coming to us. And before I knew it, I had 200 patients with red syndrome. We became the biggest center for red syndrome purely from just an observation. And that's where I decided as soon as I finished my neurology residency to help these girls, I'm going to go find the gene and figure out what to do about it. And I found a mentor, Dr. Arthur Bedetto took me into his lab. And I told him I have 200 patients with red syndrome. I want to find the gene. He was excited. He said, show me the families. And that's what they look like. One in a family. They're all sporadic. And that's when Art told me, no way. You can't work on my lab on this. You'll never find this gene. Remember, this was 1985. PCR wasn't even yet discovered, let alone the genome and so on and so forth. So I worked on spinal cerebellar attacks. Yeah. That's what I started in his lab. But never forgot that I kept working and working with the hypothesis because they're always a girl. It's probably on the X chromosome. So I focused on the X and literally marched gene by gene, have to characterize the gene to find its exon interim boundary, sequence it on the bench. We couldn't ship sequences back then. And eventually, with a tenacious postdoc in the lab, Ruthie Amir, we identified the gene in 1999. So it took a while. Today, literally, you know, if you look at the Guinness World Records, Baylor College of Medicine was part of that team. You can find a gene within five hours from drawing blood to finding the mutation. Took us 16 years. You can find it in five hours so that you can use the next 16 years to do biology. So I share that perspective to show you what an exciting time you're living in. So it's a gene that encodes a protein called methyl cytosine binding protein 2. And what is that protein? That protein was described nine years or seven years before we found that a gene was discovered biochemically by Adrian Bird in Edinburgh. And Adrian found that it's a protein that binds methylated cytosine. So you're familiar with the four bases of DNA. And those makes the genetic code. And cytosine is one of those bases that can be modified. And you're modified by adding this methyl group. So it's a methyl cytosine. And the idea is when you do that, now you can change gene expression without changing DNA sequence. And a lot of studies have been done on that. I will just update you that it used to be thought the cytosine has to be followed by a guanine when it gets methylated. But now we find and work from the Eckers lab showed that it could be any nucleotide after the cytosine. And the most common one is A. So MCA is also a mark. And that's the one we believe we contribute the most to RET syndrome. So what did we learn after finding the gene? I'm going to start first with human data, then move to mouse data, then move to the biology. RET is on the, the gene is on the X chromosome. And many of you know females have two X chromosomes. But nature's not going to get, let females get away with something that's twice as much as males. Males have one X, females have two Xs. The way the nature does it, so the males are equivalent in their gene expression levels to females, is in every cell in a female, only one of the two Xs is expressed. And the other X is silent. We call that X chromosome inactivation. So in every female here, half of her brain cells express the X from the dad and half will express the X from the mom. And in a male, it's the one X in every cell. So now males and females are equal. So in RET, the mutation 95% to 98% of the time happens on the X from the dad. A new mutation in one sperm and the dad damages this gene passed on to the daughter. And therefore, you can imagine in this RET female brain, all the cells that are empty, that don't make this protein are the cells that inherited the paternal X chromosomes and have it active. And the cells that have the healthy maternal Xs active make a protein. So the RET brain is mosaic. 50% of the cells make, express the protein and 50 lack. And think about this. Just having 50% of the cells lacking the protein, look at the disease. Regression, cognitive problems, stereotypies, autism, apraxia, this inability to plan and move, tremors. They have lots of autonomic and breathing dysfunction, seizures. As they get older, they develop stiffness and Parkinsonism. It's not lethal, but it is a very devastating disorders. The simplest way for you to remember the phenotype of RET syndrome is to say everything I can think about as a brain function, think about it being disturbed. Not normal. And add on top of the head constant hand wringing. That's RET syndrome. In males, it's very rare because remember, this comes 95% to 98% from the dad. Occasionally, a mother who happens to be lucky and she carries the mutation, but only the cells that express the wild type allele are active in her brain cells. So she has all her brain cells healthy. That mother is at risk of passing the gene to a daughter and a son. And those sons will lack the gene in every cell because they have a single X. And sadly, they're going to be very, very sick and they're going to die. By two years, they die. So you can't tolerate the loss of that gene. The beautiful thing about humans is that I call it a wonderful model organism. It can teach you so much about spectrum of clinical phenotype and mutations. We've identified mutations that are very, very mild that in a woman will have no effect when it's only in 50% of the cells. But in males, having the small mutation still makes a protein, but it functions at a lower capacity or binds a DNA less or the level is a little bit lower. Those males will always have mild learning disability, but on top of that will have a psychiatric symptom. These are in different colors. We don't see them all in the same patient. One may present with schizophrenia. One may present with bipolar, one autism and so on and so forth. So the full picture then is from zero protein lethality to very mild dysfunction, neuropsychiatric symptoms. So that's what we learned from the human patients. Now I'm going to move to the mice. The first thing we did is we made a mouse model of Rett syndrome. We took the gene out, the mouse developed features of Rett syndrome, everything, all the features we see it. But we were interesting to see if we can add the gene as a tool for gene therapy. And a genetic student in the lab, Anne Collins, made a mouse that had its own McP2 gene, but now she added the human gene. And she used a large chunk of DNA that has all parts of the gene and all its regulatory element, but nothing else. It was 90 kb, that's all it had. So it's now expressed in the right time and the right place, but it's an extra copy. So these mice make twice the normal level of McP2. This was our positive control. We thought we were going to do gene therapy and add the gene back. Let's do the ideal gene therapy, express it in the right place, the right time, see if it rescues the null mice. If it does, then we do the vector gene therapy. Well, it turns out that having twice the amount of the protein is really bad. You see these mice having a seizure, falling, having a lot of neurological abnormalities. And we wouldn't have never known that if we didn't do this mouse experiment. And this led us to ask, could there be humans with the duplication of this gene? So red is mutations that inactivate the gene, but Linda Van Esch actually was the first a year after we discovered the mouse to discover these duplication males. And this is my first duplication male who had presented with autism and gradually deteriorated. And sadly, he died by the age of 20. And these are the features. And all the features we see in the mice, every one of them we saw in the human. So this told us this is a Goldilocks protein. Too little is bad, and too much is bad. I want to bring you to a little bit more recent to give you a little bit more insight. We recently identified some of these elements that regulate the expression of this gene in the brain. And we deleted them one by one to see their effect. And there's one element, if we deleted it, we decreased the protein by 30%. And when we look in the mice, these are all the features of red. We see that a 30% reduction gives you hyperactivity, anxiety, and social deficit. On the opposite side, here's the duplication mice. They have all these features. We found another regulatory element that if we deleted, we had, again, cognitive and autism-like phenotype. So just decreasing by 30% and increasing by 50% is enough to give you a feature in the mouse, which tells you in the human it's going to be really strong. And two important lessons to learn from this. Not all neurons are equally sensitive to the protein, because the protein was reduced by 30% all over the brain. But yet all the motor and seizure problems did not happen, whereas it's mostly those neurons that are engaged with social behavior and psychiatric-like phenotypes that appear. So it tells you our brain vulnerability. As you think about the brain, that's probably why we see a lot more autism than we may be seeing new onset motor disorders, right? Because the motor system is a little bit more resilient. That's one. The second one, it tells you, if we're going to do therapy, even if we didn't reach the normal 100% level, any direction we reach, as long as we get it in all brain cells, we're going to help either disease. So I'm going to summarize to you all that we've learned from mice and humans and show you that if you lose the protein, it's lethal. If you lack it in 50% of the cells, you get threat. If you reduce the protein by 50% in all cell, you get a disease. I showed you the 30% and the same on the other side. And even tripling the protein is even more severe, cause death in infancy in human, and the mice are more severe. So now this gives you a picture, and it tells you that somewhere between 30% and 50% on either direction, they're going to be subtle reduction by probably, I predict, 10% to 15% that might cause psychiatric disorder, isolated late onset psychiatric disorders. So what does this protein do? And what can we learn about what happens to neurons? Why is it that these neurons are not working well? Well, one of the, when Adrian discovered this protein, because it bind methylated cytosine, was predicted to be a repressor, one that silences gene expression. But when we look at the gene expression from the brain of mice that either lack the protein or have twice as much of the protein, two things emerged. We see the inverse pattern. So what goes down when you lose it, goes up when you have too much of it. And what was interesting, if this was a repressor, you'll expect when you take it away, too many genes are going to go up. But what we see, a lot of genes go down. Many genes go up, but a lot go down. And when you have too much of it, too many genes are active rather than being repressed, but some go down. So this really could be for two reasons, right? It could be most of those are secondary, and those are the primary that could be. It could be that those are the primary and sick neurons. Maybe the neurons are hypoactive when you lose the protein, you make less gene expression. So most recently, it took us a while to figure out how to address this specific question, but it took me getting an engineer to do it. And that's Sameer, who did his PhD at UVA in biomedical engineering and joined the lab really to figure out this problem. How can we figure out what's primary, what's secondary, what's really happening in the neurons? What's the earliest molecular change when we lose this protein? The other intriguing thing, remember Ashley's story, I told you it took two years for her to manifest the phenotype. We've done a variety of mouse knockout. We've done a mouse knockout in the adult, and we show we reproduce red syndrome. If you just let the animals grow till they're three months old and knock it out, you get red syndrome. But just like early on, it takes months for the phenotype to appear. When we knock it out in the adult, it takes months for the phenotype to appear. Why? Why is it taking time? So these new results I'm going to share with you are to address those questions. What's the earliest and why does it take time? So how are we going to address these questions? So Sameer really did a heroic study. Here's what he's done. He grew mice till they were adult. These are mice with a conditional allele for MacP2, which means you can delete the gene at any time you wish, in any cell you wish. Here he deleted it in every cell, but in four months old mice, so in adult. And the idea was if you deleted at 16 weeks, you wait a week and you look at the RNA, you look at all gene expression changes, you look at histone marks, study physiology, behavior at that age. Then come with a new cohort, do the same, but now look at two weeks, same three weeks. So basically think of it as he's done six different knockouts. And in those six different knockouts, you look at the gene expression, chromatin marks, physiology, behavior, etc., etc. And this is what we learned. Here's healthy animals. They make a healthy dose of the MacP2 RNA. The RNA within one week is gone. So the RNA goes right away. The protein lags a little bit behind. As you can see here, it's only reduced by 30%, and by 50% around nine days and so on. It's finally gone by four weeks, but it's reduced. So that tells you this protein has a long half-life, which is not surprising. It's embedded in chromatin, so it's not surprising. It hangs around long. But we started looking at gene expression at all of these time points. Even when there is 30% reduction of the protein, are we going to see gene expression changes? And the answer is yes. You will see here when only 30% is gone, we start seeing some. When 70% is gone, we see some, and so on and so forth. And one important thing to say is some of these things that change, they continue to change over time, and they get bigger in magnitude. And then we took these gene expression changes at eight weeks of age that we saw many of them early on and compare them to the mouse that lacked this gene throughout its life. We called it the constitutive knockout. And what you see here, very nice correlation. Here's the correlation coefficient, quite a bit of nice correlation. So this told us that really there is not a developmental phenotype to this gene, because we're finding strong correlation with the adult knockout. So there's nothing happening as the brain is developing that we're perturbing by losing this gene. That's important. And it told us that many genes are altered early on. And then he did what we call cut and run for us to find where this protein sits. And you'll see here, I'm just going to show you that here's red is in the knockout at one week, two weeks, four and eight. And black is the wildtip. And you'll see the protein binding throughout the genome. This is one chromosome is slightly getting reduced till it's almost gone. But it's reduced everywhere the same. It's not like we're finding more changes at those sites where the genes are changing. We have not seen that. So we're trying to figure out what's driving the gene expression change at those sites. So Samir did other chromatin mark, other histone marks that really either denote activation or repression of gene. And he did four or five of these. And I'm just going to show you one of them. This is a mark that is a mark of activated gene expression. And what you'll see at one week, we're already seeing at those genes that are downregulated significant reduction in this mark. For those that are increased, we start seeing it at four weeks. So this tells us really the reason we're seeing those up and down genes, for example, is this reduction in this mark that activates genes that's not happening. He then looked at physiology. And at four weeks, this is synaptic plasticity assay in the hippocampus. You'll see that this happens after gene expression and chromatin changes. Perfectly normal. But at eight weeks, we start seeing a change in plasticity. And we now did another experiment at six weeks. And this experiment showed us that it is intermediate. It's not as prominent as the eight weeks, but it just begins to start. So I'm not going to show you physiology and behavior on eight knockouts. I'm just going to summarize it in one slide to tell you chromatin changes are the first thing that happened. And gene expression changes are the first thing that happened. There are some behaviors that happens before others. The body weight changes earlier. And we know this is from the hypothalamus. All our physiology was hippocampus, and a lot of our behavior was hippocampus. And you'll see that happens later. So we believe it's these molecular changes that are the root of this change. And these data really, we haven't published this yet, but I'll be happy to share those with any of you engineers who'd like to really look at patterns and think of what's driving this change. So let me just show you. So what did we learn from these studies? We learned that the gene sensitive to the loss of this protein are the same, whether it's adult or later. And that the changes in chromatin, it's the one that accompanies the rapid depletion of the protein and gene expression. And these gene expression changes are primary. They're not secondary to abnormal circuit activity. So what's happening in red right now, we and others are trying to discover a way to manipulate the disease. And there are efforts in gene therapy. But one of the things that we learn from these studies is the dosage is important. So some of these gene therapy took our animal model with that application consideration and tried to do gene therapy by regulating the protein they put in. So it's always in check. I would like to now switch to the duplication because in our lab we felt it was easier to tackle the duplication therapeutically than a protein missing in half of the cells. And for that, we wondered if you normalize the level of the protein, we did it first genetically by waiting till the animals are adult and take out the gene and all the symptoms reverse. The next one we asked, can we do it using a small molecule, an anti-sense oligo nucleotide? Those are small pieces of DNA that will bind the RNA. And when you have DNA RNA heterodouplates, an enzyme in the cells RNAs H will degrade it, will degrade the RNA. So an anti-sense is a way to lower the RNA. And Hessie Steinberg and Yingyao Shao did these studies. We collaborated with Ionis pharmaceuticals. They made us these oligos against McP2. And you could see here one example, this is normal mouse activity. You could see these duplication mice have decreased activity. But if as adults, we treat them with ASO, we reverse the symptoms. Their activity is normal. And this is a group of animals. This is one animal. It's nice when you see the phenotype based on one animal, but this is a group of animals. All the symptoms here are corrected with the ASO. This was done at about four months of age. I wanted to take the latest time possible. I could study these mice, age them as much as possible, and see, can I still reverse the symptoms? Well, the oldest this mice can live is about nine months. They start dying at nine months because they have seizures all day long. And having seizures all the time, they eventually die. Sadly, that's also seizures contribute to the death of the patients. So we waited till the mice were that old and then gave them the ASO treatment and see what happens. And here's just an example of the EEG, electroencephalogram that measures brain waves. You see the excitability because of seizures. You see it now normalized with that treatment. And all the seizures stop. So this was really exciting to us. It told us at least maybe even in nine months, all the infants would get reversed it. But the question, could we do it later? Because nine months old mice don't equate necessarily nine months old infant. They may be older. However, we quickly learned, humbled that not so easy. We published this paper. We thought, that's it. We're ready to go to therapy. But Ionis told us, which is correct. Remember, the humans have two identical genes. In your mice, you had one human gene, one mouse gene. You eliminated the human. You safely still have the mouse gene. The humans have two human genes. You want to make sure you don't go too low to give them red syndrome. So we created a new mouse model. I repeated all the studies and again showed you can titrate this drug just like you would titrate a drug for hypertension, for blood pressure, or for blood glucose, right? If you give too much, you're going to put the patient in coma and they're going to pass out. So it was the same way. And I'll summarize those studies. Essentially, by repeating the study in humanized mice, we were able to we were able to basically show that reverse the symptoms. And we saw that the RNA changes first, the protein next, and the target first, behavior last, just like I showed you with the knockout when you correct the same. So this gives you a molecular signature to do so that you can actually figure out which markers you can measure and say, okay, I treat it with just the right dose of ASO, or I need to give a little bit more the next time, or I need to decrease the dose, or antagonize the ASO to reverse from going too low. So this is this was a great study. I'm happy to report clinical trial readiness have been completed in natural history studies ongoing and preparing now to a clinical trial study starting next year. So this was really pretty exciting to see 10 years of work to take us from the first mouse to a potential treatment. I want to now go back to Rhett and share some things with you that you an engineer might contemplate better solutions for than I did. So we had knocked out this gene in different brain regions to see which symptoms belong to which type of neurons. And I'm going to just summarize and tell you that no matter where you knock out the gene, the neurons are compromised. That's why I said think of any brain function and think of it compromised. But once theme emerged is that losing the gene in a neuron partially disables the neuron. It doesn't totally shut down the neuron. But if the neuron functions at 100% capacity, it now functions at 70% capacity. So you've got all the neurons, but they're just functioning less than normal. So and we couldn't rescue if we rescue in one cell type, the animals will get better for a little bit and then they crash. So we decided this is really a circuit problem. And we should look at the circuit. And Lindsey, he decided to look at the hippocampal circuit and watch the neurons during a behavioral paradigm. So she put this calcium sensitive fluorescent protein where when the animal is moving, you can now see these neurons fire and see how do the red neurons look when they are learning. So this is really capturing any live behaving animal, how its circuit behaves when it's trying to learn to see why do the red mice don't learn as well as the healthy mice. And you can see here we reproduce that these mice, they can learn within an hour. I'm not going to go through the test to a type test, but you'll see they're different at one day. They cannot remember, they got shocked in this cage. They don't remember that. And that's different. And she looked at the network to see what's different. And what you'll see here while in wild type animals, when they're learning a few cells become dedicated to that activity, the red neurons, all of them are active at the same time. They don't really differentiate. They don't firm what we call an engram to say, this is the task I was involved in. So if I was to put this in a cartoon form, you'll see that in a healthy mouse, when you teach it a task, a few neurons become active. But in red, all of them are functioning at the same time. All of them are active at the same time. There's not that selectivity that helps them. And we were able to then dissect the circuit and find that it is really the lack of inhibition. So the way you get those selection is by inhibitory neuron coming and inhibiting everything around so that you have a select group. And this was not happening. And we were able to show that. So having learned that in the red brain, everything is functioning as reduced capacity, we said, why don't we stimulate the red brain? And to do that, we collaborated with Jean Rontang in my institute and his post-Axuang Hao. They did deep brain stimulation in the part of the brain called the fornix that project to the hippocampus where this test I just showed you for learning and memory is happening. So you plant electrodes in the fornix. This shows you where the electrode was. And when you stimulate, you record from the gyrus, dentate gyrus, and you see the right activity. I'm not going to go through the whole experiment, but they use the same paradigms that neurosurgeons use for patients with Parkinson's disease or other movement disorders that have been treated by deep brain stimulation. And I'll summarize here that deep brain stimulation corrected all the hippocampal phenotypes, all the learning and memory, everything was corrected. So the gene expression studies were corrected. And the selection of neurons that I showed you wasn't happening now is happening because now the inhibitory neurons became more active. So this is not a solution because red affects the whole brain. Are we going to do deep brain stimulation? Every brain cell is going to be impossible. You can do at most two transducers. And we might. We now show data that it helps motor functions. You can do maybe motor and learning. But we wonder, how is it doing it? And could we substitute for that intensive deep brain stimulation with perhaps training on a task excessively? Because the more you learn something, the more your brain cells become plastic and you do it better. But we knew parents do this with physical therapy. So we suspected it's only going to work a little bit. But could it be if we started early before symptom onset? Could we do better? And this is what Nate Achilles did, the graduate student who did his work. He put the mice on a rotating rod. The rat mice usually fall. If you look at these mice, they start falling at 12 weeks of age. By 24 weeks, they cannot stay in the rod. So what he did, he took them before their symptoms onset, just like the human, rat mice, don't develop symptoms right away. And he trained them every two weeks. And they got 18 sessions of training. And he compared what happened if I trained before symptom onset or I bunch it all up after symptom onset or no training. And the results were obvious. You could see untrained, they fall right away. Late training helps a little bit. But really, it's the early training that makes a difference. And that was really nice. And we followed the mice for months. We delayed the onset of motor incoordination by six months. And even then when it happened, it was very small. So it's really exciting. Then we asked with this work for another part of the brain. So we checked the hippocampus where you put a mouse in a pool and it's going to hate being in water. This is hidden, but it's going to use the cues on the wall to swim and learn that right near the star, there's an escape that platform. That's how they learn. And if you do this with healthy mice in four days, they will learn. But if you do this with rat mice, they don't learn that task. And what you see here, we had to do this test a little bit earlier because this test developed earlier in symptoms. And again, we did the same thing a couple of sessions once a month. So it wasn't intensive training, but it was enough. And you'll see what happens. Here's healthy mice in four days they learn in rat. The red naive mice, they don't learn. If you do late training, they don't learn. But if you do early training, they learn as well as the healthy mice. So we then challenge them. We remove the platform and we see, do they search for it? And you can see that the rat mice that have been early trained, but not the naive or late, will search for it, will cross the area and will do so much better. So I think, do I have five more minutes or not? Where is my timekeeper? I have five minutes. Okay. It was interesting. We were very excited about this result and thought, this is it. We're just going to fix it by training. But we discovered quickly that the mice did great on this test, but not on a different hippocampal test. The electric D-brain stimulation hit every neuron. But this training was only engaging the neuron engaged in the task. So we only got task-specific improvement. And this led us to ask this question, is it task-specific? So we designed a strategy to label and trap those neurons. You can see here, these are the neurons that are activated in the Morris water maze. If you handle the mice, nothing. And now we ask, if it's only those neurons engaged in the task, if you silence them, you should lose the effect. And if you selectively activate them without the swimming training, you should gain the effect. And for this, we used Brian Roth designer receptors that are activated by CNO. We express either the one that silences or the one that stimulates. And you'll see here, when we silenced those receptors, now everything that you can see here, we replicated the experiment. The training, they perform as well as wild type. But if after this training, we just silenced those neurons that were activated by learning the task, we lose the effect. In contrast, if we put them once in the pool, and then we don't anymore train them, but we simply activate those neurons by the activating receptors, simply the same neurons that we trapped, we activate them, they will learn. And we showed the same, that the physiology improves. So I want to summarize here and tell you what we learned from these studies is early training does really work, but it's task-specific. This is clinically important for two reasons. If we can start diagnosing at birth, like doing neonatal screening, we've got about three years, because that's when the typical diagnosis is made molecularly. We've got about three years of training that we could do these infants. And that might delay their onset of symptoms, perhaps by months, perhaps by years, and make them more receptive to more definitive therapies and enhance their functionality. And it's something that we're contemplating doing. The other thing it told us, because of the way the red brain works, so many neurons active at the same time, to get the benefit, you cannot do physical therapy on three different tasks. You have to train them one task at a time. So I've been now informing physical therapists how to train red girl. And I have one newborn I'm following, and she's seven months old, and she's looking advanced and our fingers are crossed for her. So in summary, I just want to thank, I mentioned all the people who've done the work. Sameer asked me to tell you he's starting his own lab at UVA. So if any of you are interested, he's working in autism. And my great thanks to the families and the patients who participated in the studies, our funders, and our collaborator, Jean Denglu, has been our computational biologist. I mentioned Jean Ronk, and Matt helped us with the network studies. And this is a picture of the lab. There's never one picture that captures everybody at the same time. So I always have to do different pictures because they graduate and I'm super proud of them. And this is the institute. Thank you. So we have a lot of questions. I'd like to give the honor to the students in the audience. Any student want to take the first shot? Hello. It was very nice. I really enjoyed it. So my question was, if red syndrome is also affecting motor skills, which is the long range gene. And I am assuming death is because of the motor skills, like eventually the heart muscles and everything would give up. And that's how death is happening. But are you also looking at the motor skills, which are so long range? So it does affect motor skills because it affects cortical motor neurons, spinal cord motor neurons. It does not affect the heart. It only affects the heart through the nervous system control. Some of the sodium channels that control rhythm in the heart are affected in the brain. Adrian Byrd did a very nice study where he compared the brain only effect versus the rest of the body. And basically all the red syndrome phenotype are purely from a nervous system origin. I cannot, Glea will probably contribute a bit, but it's nervous system origin. Okay. So there wouldn't be any dystrophy of any other muscles? No, the muscles are okay. It's really more the neurons that control the muscles. Thank you. Thank you so much for that really interesting talk. I had two questions. So first, since there's a degree of mosaicism with red syndrome, is there a sort of spectrum of how patients present in terms of severity? Yes. So that's the first question. Typically classic red syndrome is when you've got the mutation expressed in 50% of the cells. And if you got the mutation that's inactivate the protein, it's very severe. Two things that may slowly decrease their severity. One, if sometimes some cells with the healthy protein tend to proliferate better. So the brain is not 50-50. It's 70% the healthy X and 30% the mutant X. Sometimes it's 80-20. When you get that, you get milder symptom. You may get learning disability and psychiatric symptoms rather than getting full blonde red. So that's one contributor to the severity. Another contributor to the severity is the type of the mutation. There are mutations that totally abolish the binding to the DNA or delete the gene. These are very severe. But sometimes you get a mutation at the very end of the protein that simply lowers the level of the protein by 40%. Those are also milder mutations. So they have a partial phenotype. They may say a few words. They may continue to be very ambulatory. But they'll still have multiple features. Okay. That makes sense. Thank you. And then my other question. Do you know a couple of examples of maybe the types of tasks that physical therapists would have, especially really young patients doing? That's a great question. People tell me, how are you going to train a newborn? And the answer is easy. For those of you who have kids, you remember when you were sent home with your newborn, they told you, make sure you do tummy time to increase their core strength. You do tummy time two or three times a day, five minutes. You do it for at 10 minutes, maybe four times a day. You start strengthening their core, helping them get up from a lying position. So you can strengthen their motor activity early on. Typically, a one-year-old parent will say, dada, mama, trying to teach them a word, in passing. For red, you've got to be more intentional. You sit with the infant and you keep saying, mama, you know, repeat, repeat, repeat. It's just a matter of focusing on one thing at a time. You're not going to do the mama and a cat at the same time. The red brain can't take that. So I think this is how we have it. It's difficult, but I think it's doable if you focus. And you don't have to do it too many times. You saw we did it every two weeks in the mice to get a difference. So here you think just add an extra training. Thank you. There's a question here. Do you know if there are any papers which talk about like bimanual coordination in these kids early on? What kind of coordination? Bimanual coordination, like training bilateral integration. Their bimanual coordination is really good early on. As you could so, Ashley was holding a book using both of her hands. She could hold on the horse, you know, using both of her hands. So early on, there's nothing. It's really after the cells sense the absence of this protein for a while. Yeah. Yes? That happened and lead to delayed behavioral symptoms. But do you have any hypotheses in the child? They're born without it. And yet you still have a relatively long delay. That's a great question. We believe the long delay is because of the epigenetic marks. So remember, this is a protein that read methylated cytosine. Initially, we all believed it's reading methyl Cg. And that's what's driving the disease, which is there in the brain from birth and throughout. It turns out the one that's really important for red syndrome, Laura Lavery did that work in our lab, is the methyl cytosine A followed by adenosine. And MCA is not written early on in life. DNMT3A adds this mark after birth. So it adds it after birth, typically in mice, five weeks. And that's about the time we start seeing symptoms, six weeks to start seeing symptoms. And in humans, it's continued to be added till about young adulthood. But the majority of it is in the first two years of life. So it's because of that. The protein is not needed for neuronal function. Neurons have to mature. And then the mark is laid on. And then the protein, this is the reader of that mark. Thank you. Great questions. All right. Of course, I'll say it was a great talk as well. Thank you very much. So I'm wondering if you think in terms of a personalized approach to red syndrome, thinking about the mosaicism, if you think about the training paradigms you mentioned, or even the chemogenetic approaches that would mimic those. Would you think about looking at neurosurgery signatures that would help guide specifically tailored types of approaches, either behavioral or chemogenetic? Right. Thank you so much for asking the question because I wanted to amplify where I think everyone in this room can really make a meaningful contribution. So we're scratching our head. Of course, one day when we have a gene therapy, right now there are gene therapy trials, they're being started, they're going to hit 15% of the cells. And we're hoping they really, you know, put the protein in the missing cells and don't put too much in the wild type cells not to create disease in those. So it's a start, but we need other pathways. And one of those pathways, I mean, to me, the most dramatic rescue I've ever seen is from the D-brain stimulation. The training is good, but you only have a window to do it. So what's an alternative approach, stimulation, but can we stimulate more of the brain than just, you know, this region or that region? And how do we do it so that we stimulate and train and go back and forth, reiterate, like put the stimulation together with the physical therapy training? And this is where perhaps improvement of transmagnetic stimulation, this is where the engineering will come in. I think what we really need is some sort of, I can imagine, and you will do it, I'll just give you the ideas. I can imagine some surface electrodes that could be modeled and tested initially in, say, non-human primate to say, if I stimulate this portion, I'm really hitting the hippocampus right here. If I'm doing this, I'm hitting the cerebellum, so that you do it while you're training and you do it gently. And then you keep it throughout at a very mild level because I told you all these neurons disabled. There's two types of responses. You want a response that they're never so dull, so that you do low level throughout, but you want a response when engaged in activity. And typically when a neuron engaged in activity, like I pull something, those neurons controlling that are going to be more active. So I think we need to tweak this, and this is all engineering problem. That would be my preferred approach, an electrical engineering problem, and I really think things like this can happen now. I mean, I'm seeing exciting things from some of our neurosurgeons they're doing. The alternative is eventually using molecules like the dreads, but safer, CNO is toxic, where you sort of hope to give these and test activation at will. But that's hard to be regional in a human, whereas electrical can, too. We have two. Thank you so much for wonderful talk, actually. Open the eye for me for all these knowledge. Okay, my one quick question is that in your rescue experiment, particularly it's a deeper stimulation. You have a rescue restore some of the function. At those restoration happen, what's the molecular base of these rescues? Are they in the gene level, protein level, or circuit level? Thank you for asking. Yes, we did all these studies, and when we did the stimulation, we rescued the gene expression. We normalized 25% of the genes back to normal level and another group slightly, and the ones that were normalized, it's really interesting. They're all involved in neuronal function, synaptic function, and they're typically misregulated in other intellectual disabilities, autism-like disorders. So these may be a set of genes that are really sensitive. If they go down, you start seeing intellectual disability and cognitive disorder. So that was one of the first things that was rescued. It was simultaneous. My bet, given the adult knockout and the cascade of the disease, my bet it was molecular first physiology next. We saw rescue of both. LTP normalized, in vivo LTP, neurogenesis normalized. So we see all of the above. We really rescued first gene expression, second all the physiology and the behavior. And I thought you're going to ask me, how durable is the rescue? So we did the treatment for two weeks, and the rescue lasted for about 10 weeks. So it was pretty good rescue. So you can do it, you know, per time. You don't have to do it constantly. Thank you for this great talk. It's as clear and impactful as the first time I heard in Atlanta, Georgia. So that's my comments. My question is, the red signal is mainly a neuron developmental disorder, but it may also involve some neuron degeneration later on, right? Oh, there's no such thing. So can you comment on this part? Great question. It doesn't really involve neurodegeneration. So we've done a lot of extensive pathology in my old days. When I was still a clinician, I worked with neuropathologists. A small percentage of red girls will die prematurely from, I told you, their sodium channel genes and other genes that control heart rhythm. There's a slight vulnerability in heart rhythm. They have abnormal, corrected, long QT, QT interval. So they have premature death. We think it's cardiac. We don't know. But a small percentage of that. So we've looked into brain from different ages, and all the cells are there. There is no degeneration. The only reason the brain gets smaller is because the dendrites and number of synapses decrease. And this is the, unfortunately, I didn't show that one slide. It's the one of the slides I eliminated. When we did the training, we increased the dendrites. We increased the number of synapses. And nuclear size shrink with the loss of the red syndrome gene. Again, if you reverse it in the adult, you correct that. Great. Thank you.