 So, it's a real pleasure for me today to introduce to you today's Steenback lecture, Huda Zagbi, who I thought I would give a couple of minutes of introduction for Huda. She comes as an immigrant to the U.S. and was one of our star immigrants. I'd like to point that out. She came from Lebanon. And in Lebanon she started her medical school there and then kind of halfway through her first year of medical school, the civil war broke out and that was not good. She came to the U.S. to finish medical school, then went to Baylor. And the rest is history. She's been a professor at Baylor for, I don't know, a long time. Not as long as I've been here though. She is a Howard Hughes Medical Institute investigator. She's a member of the National Academy of Sciences. And the founding director of the Jan and Dan Duncan Neurological Research Institute at the Texas Children's Hospital. So, she's got lots of hats that she wears. I think one of the things that's quite amazing about her over her career is how she's an MD who has focused on trying to understand disease and the molecular basis of that disease and managed to go from phenotype all the way into molecular mechanism and basic science in a beautiful way. So she really spans that breadth of basic science to disease. And I think it's particularly suitable that she is here as the Steenbach lecturer because Harry Steenbach was somebody who did something rather similar. He has a picture of Harry Steenbach who was born in 1886. He's a little older. He got his undergrad degree in 1908 at the University of Wisconsin-Madison. He then stayed here for graduate school and he was, I think, started as assistant professor in 1916 and then graduated through the ranks by 1920 he was a full professor. So things were a little different in those days. And his claim to fame, which is a rather big one, is that he also spanned the basic science biochemistry to disease. And he discovered that UV irradiation of food generates vitamin D and very importantly he patented that, which is why we can have these lectures today. He not only patented it but he, and he spent $300 of his own money to do that because the university wasn't willing to do it. And then when Quaker Oates wanted to buy his patent, he decided that it was smarter if he got a consortium together and started a licensing company at the UW, which is now Wisconsin Alumni Research Foundation, or WARF. And he got nine of his friends together, he and eight other friends. They each put in $100. So WARF started with a $900 initial budget and it's clearly grown since then. But the patents have been really important for Wisconsin. And again, he's done exactly what you've done except in his own era, which is going from the very sort of disease-oriented rickets. He basically solved rickets to the molecular basis. So with that, I'd like to also just mention that I consider Huda a rock star of molecular disease and she won the 2017 Breakthrough Prize in Life Sciences for her work that she'll be telling you about today and tomorrow. So welcome, Huda, it's great to see you here. Thank you so much, Judith, for inviting me and for giving me the honor of giving this T-Box lecture. It's really, he was an amazing man. And to walk today through what the WARF has built and look at the various research initiatives here, it speaks to the power of basic research and the stellar biochemical research in this institution. So I'm really honored to give this lecture and thank you for organizing such a fabulous program. Everybody I've met with, it's been very exciting. So today I'm going to tell you a story, two stories actually that are put together mostly to sort of inspire you to think about how neurodevelopmental disorders will come about and what does it really mean for the broader class of neuropsychiatric diseases. So this is the way I prepare the stories today. I would like to disclose some affiliations. The work I'll be presenting tomorrow will involve collaborations with UCB Pharmaceutical given basic research we've done, they've taken that to develop potential therapeutics and that's relevant. Some of the work I'll present today is in collaboration with IONIS Pharmaceutical. We do not receive any money from IONIS but they do provide us the anti-sense oligos and they also collaborate with my collaborator Harry Orr on attacks in one and then I serve on an advisory board of Denali and the board of directors of Regeneron. So the first story I'd like to tell you is the reason I am actually a scientist. It's about Rett syndrome and I was inspired to work on Rett syndrome by actually the patients I saw particularly Ashley was the first patient I saw with this disorder and here she is at five years of age about the time when I saw her in the clinic when I was still a first year neurology resident. What's striking about Rett syndrome is that girls like Ashley are born healthy and then go through a year to a year and a half of normal development doing everything a typical girl would do but gradually would lose all the skills they acquired. Their head growth will slow down and they will lose language communication, social communication, have balance problems and seizures and breathing and many autonomic problems. Essentially every part of the nervous system is affected and when you see a child with this disorder it doesn't leave you and I happened to see Ashley she was actually at the time there were no cases diagnosed in the United States there was only one report about European cases and seeing her leaves an impression on you. How devastating it is to really see someone healthy and then losing everything they've learned how to do. And it just happened so ridiculously that I saw a child a week later with the same symptoms. So seeing two children with a disease no one has ever seen in my institutions or at that time in the United States as far as we know was really left an impression on me and inspired me to look for more cases and we found them in our clinic and that's what inspired me to work on Rett syndrome and really pursue the cause of it and I'll tell you a little bit more about that. This is a video to show you Rett syndrome in action. You'll see this girl wringing her hands constantly you're going to see some rocking activity some features that people typically see with people with autism and you'll see here is she trying to walk she has difficulty initiating movements there's lots of problems with motor planning instability and Parkinsonian features. So the reason I wanted to figure out this disorder is because they were all girls and I figured anything that looks so identical from girl to girl has to be genetic and we need to find a gene for this disorder. The problem was this was 1983 when I saw my first patient in 1985 when I completed my clinical training and decided to join Rett's lab to learn molecular biology in 1985 you could not find the cause of a sporadic disease and all Rett girls were typically like this one in a family. So this of course raises doubt whether the disease is genetic but more importantly it was not possible with the technology available and this is the reason I worked on the ataxias. That was the project that Art told me you have to pick a project to learn molecular biology on how to work on the ataxia while trying to figure out what to do about Rett's and you'll hear about that tomorrow. Fortunately with perseverance we found a gene in 1999 so 16 years after meeting Ashley and we discovered it to be a gene on the X chromosome which explained why they are females and it encodes methyl CPG binding protein 2. At the time the gene was not associated with any disease but about 9 years before we discovered that Adrian Bird has found it as a protein that bind methylated cytosine and as you see in this cartoon. So that's what we knew about it at the time and it still is. It's a protein that bind methylated cytosine. Now we know not only CPGs but also Cs followed by any nucleotide. So what I'd like to talk to you about today is tell you what we've learned since the discovery of the gene. What have we learned about the phenotypic spectrum of my P2 mutations and what are the effects of losing this protein on neurons? How can we treat this disorder? What therapeutic strategies do we have? And then what I have learned from studying this one rare disease about 1 in 10,000 people and how it is relevant to the broader neuropsychiatric disorders. So on the phenotype I mentioned to you what classic red syndrome girls look like and basically these are typical symptoms and we call that typical or classic red. What we discovered although all the people with red were girls we learned after the gene discovery that males who have a mutation and because this is on the X chromosome they don't have the protection from the wild type cells and they are typically very severe. They have encephalopathy, they have motor problems and sadly when they have a mutation that totally inactivate the protein they will die within one to two years of life. These are the girls survive and do well and the oldest one can live into their 70s and 80s is that they're mosaic. This is on the X chromosome so only one of their axis carries the mutant allele, expresses the mutant allele and the other axis typically express the gene with a healthy allele. What we also learned though is that if a male has a milder mutation or a female for this matter but mostly males if they have a milder mutation which just partially inactivates the protein then we see individuals that may present with totally different features and most of these are psychiatric features, autism, anxiety, obsessive-compulsive disorder, attention deficits, bipolar and schizophrenia. Each is in a different color because the same patient will not have all these features. A patient may have just anxiety, hyperactivity and tremors or one may have bipolar and spasticity and so on. They all have mild learning disability so this is in this color because every one of them will have mild learning disability but the psychiatric features are different. So why is this important? It's important because it tells us that the milder the mutation the more you see phenotypes in the psychiatric spectrum which tells you that these are probably the most vulnerable circuits and that your mutation is so mild that you're not seeing as much motor deficit perhaps but you're seeing psychiatric features. So this is in the humans and in mice we could model this disease and the mice have all the features of the disease. So here are the features of the disease and when you mutate this gene in mice you pretty much have almost all the features and this led us to ask the question what happens if we delete this from particular neurons in the nervous system would we get a subset of these features? Is there one particular neuronal type that's really critical because that will tell us maybe what neurotransmitter systems are most perturbed and maybe we can do something about them. And here I'm just going to summarize many decades of work and the work of many postdocs with a couple of examples. These are examples from work by graduate student Shanling at the time graduate student Tuan and two postdoctoral fellows that studied the effect of loss of this gene and either the excitatory neurons in the brain or the inhibitory neurons. Now excitatory neurons make up about 80% of the brain cells in neurons and inhibitory neurons the other about 20% and they come in different subtypes. So we studied the effect of loss of this gene in either type and here's what we learned. What we learned is that if you take it out from excitatory neurons you get anxiety-like behaviors, tremors, and obesity, problems with sensory motor gating, motor coordination and the animals will die prematurely. Those same problems occurred when you take it out from inhibitory neurons. So what this told us that this protein is essential for the function for neurons that control these behaviors. But what was interesting is that the inhibitory neuronal loss manifested all the other features of the syndrome, the repetitive behavior that you saw in the girls, the hand-ringing which we see in the mice with forpo activity, the seizures, spasticity, the inability to perform motor functions properly and social problems, all of these we saw them with the inhibitory, but not the anxiety or the tremor, those were excitatory. So what did we also learn from these studies? And again I'm summarizing because a lot of this published. What we learned from these studies is when you take this gene out of either inhibitory or excitatory neuron you're partially disabling these neurons. You're not totally compromising your activity, you're decreasing their activity by perhaps 20 to 30% at most. So think about it, partial inhibition of these neurons is causing at least in males all these features including death, but it tells you that different neurons have different manifest different phenotypes based on the vulnerability. Eventually I'm sure if we compromise them more we're going to see so much more features because the network will be affected. Then we divided the inhibitory neurons in the various subtypes and here again we saw this modularity. We saw that with the somatostatin neurons, repetitive behavior and seizures whereas the parvalbumin neurons that's when we saw the problems of social and motor and learning and memory. So this to us was surprising in that one we learned we're partially disabling neuron. Two, we've learned that there's modularity and it tells you which symptoms you can begin to attribute partial dysfunction of a somatostatin neuron may make someone susceptible to seizure if it's a very mild dysfunction. Because think about this now, not only in the context of red syndrome, think about it in the context. This is what happens when you partially disable these type of neurons. So to summarize what we've learned from these studies we learned that moderate reduction in GABA and glutamate can cause several neuropsychiatric features in early lethality and what this predicts that a much more subtle reduction reduction in the realm perhaps of 10% rather than 20-30% we saw when we took out Mach-P2 out of these cells might cause isolated autism or other psychiatric phenotypes and from a therapy point of view this told us and we tested that, again I'm summarizing here you can't really boost just one side of the system you pretty much have to regulate all the activity of all cell types and that means all the multiple neurotransmitters that are altered have to be you know you have to use a cocktail to treat the people. What we do now we do give them Aldopa for the Parkinsonian feature and we do give them sometimes serotonin-reaptic inhibitors for some of the anxiety features but there is really no treatment that helps with all the symptoms and this is what led us to explore other approaches until we find better treatment we decided perhaps we can explore circuit manipulation and in particular deep brain stimulation which has been used in other disorders the idea would be if we focus on one key phenotype perhaps the learning and memory and one anatomic region the hippocampus and manipulate the network activity could that bypass the deficits we see from the dysfunction of these neurons and for this we collaborated with General Tang who is the director of the M.V. of Physiology Corps and Shuang Hao is the person the postdoc has performed all that work in his group so the idea was to use the same paradigm of deep brain stimulation that's used in human for Parkinsonism but a different brain region and in this area is the forenecks which has projections to the hippocampus and so the stimulation is happening in the diagonal band in the forenecks and the recording are in the dentate gyrus so you ensure that you're doing proper stimulation you're not inducing any abnormality seizures as such and the paradigm involved taking animals that are about two months of age and then treating them for a couple of weeks for one hour a day and then examining them about six weeks later to see have they benefited from that treatment and we use the paradigms that are typically used in patients that are undergoing deep brain stimulation by neurosurgeons and in this case when we tested all the hippocampal learning and memory paradigms we saw improvement I'm showing you one example here where a mouse is put a cage and it hears a sound it gets a foot chuck and it will feel that foot chuck when it's brought back to the same cage the hippocampal learning will show that this animal will freeze it knows this context was bad it got the chuck also if it hears the sound it will also freeze because that's amygdala mediated q-dependent and in red syndrome we know that the contextual fear is altered so you see the wild type animals they freeze quite a bit because the red animals will freeze much less however after deep brain stimulation both will improve but what's really nice is that the wild type animals now will become very similar to I mean the red animals after stimulation will become very similar to wild type animals and we've done this for multiple behaviors for the hippocampus and we see this effect we also saw that in the red mice there is a decrease in neurogenesis this is the dentae gyrus where new neurons are born and you'll see I'll show you the quantitative data there is reduction of that and few of them incorporate in the network but after deep brain stimulation I hope you can appreciate that there is enhanced neurogenesis and if you look at this here again the red mice have remarkably decreased neurogenesis compared to the wild type animal compared to the deep brain stimulation that's improved in the number of neurons that actually incorporated in the network improved so now that we know that this will work this opens up so many questions what does this do to the network what does this stimulation do and this is why Wilu who was a postdoc at the time in the lab helped with MSTP student Ryan Ash in Smirnakis lab at the CA1 neurons that are the beneficiary of the stimulation and when we looked in slices using a calcium reporter you can see here in these neurons that there is some activity I hope you can see that it's sparse activity but when we look in the mutant mice you'll see a lot more neurons fighting together now this is in animals that are early stage of the disease we have seizures so this is not due to seizure activity it's just more neurons firing together and if we are to quantify that you'll see it here each of these lines is in neuron firing and you'll see in the wild type animal it's random in the null males you'll see some neurons firing together but in the females we see a lot more neurons firing together and this abnormal coincidence firing you don't want that at a baseline because we propose that that will interfere with learning and we're doing the follow-up study to show that right now and just because this is looking at imaging and slices we recently confirmed that Sarah Kea, a graduate student in the lab working with Dion G did tetroderecording she recorded from awake behaving animals and showed the same but you see a lot more coincidence in the firing in the rat mice so what drives this? here is the circuit these are the CA1 neurons and these neurons receive input from CA3 and they synapse on a group of inhibitory neurons in this orient's layer and the idea would be that these pyramidal cells that are excitatory will synapse on the inhibitory neurons which then provide feedback inhibition and it's that feedback inhibition that will keep cells from firing together so I'm sharing all this detail with you because we looked at all the synapses in the hippocampus and this was the one synapse where we found the abnormality what we found that there's decreased excitatory input onto these inhibitory neurons as measured by frequency and amplitude so now we have a synaptic phenotype that we know in this circuit is altered and when we ask what does DBS do and here's again you see the increased synchrony the D-brain stimulation restored the level of activity to normal after treatment in the mice and then when we go back and now do the same analysis with the physiology looking at MeCP2 positive and MeCP2 negative cells in the female mice, we can now label them you will see again there's decreased frequency and amplitude of this excitatory input in the MeCP2 null cells in female mice but not in the wild type cells but this is normalized upon D-brain stimulation so we have an idea how the D-brain stimulation is helping perhaps the learning and memory we now know that at least one way it's helping is restoring the firing patterns and the synaptic connectivity and what we're doing now we're proving that we're doing additional in vivo studies to prove that these oriental layers neurons are really critical for capturing that learning and memory and mediating the deficits MeCP2 as I mentioned to you binds methylated cytosines exactly what it does after that is really hard because it binds very broadly in the brain and it binds any methylated cytosine essentially it can so there's no specific targets of this protein when we lose it we find a lot of gene expression changes so we wanted to ask what happened to these gene expression changes after D-brain stimulation so Amy Pohodesh did the same experiments as you've seen before in both male and female mice and then harvested tissue to see what happens to gene expression after D-brain stimulation both in animals that have the protein and animals that completely lack the protein and there are hundreds of genes that are misregulated in the MeCP2 novel dentate gyroset baseline and some of these are up some of down, majority of genes are actually down and there are also different isoforms, splice isoforms that are changed what we discovered is that D-brain stimulation rescued the expression of about quarter of these genes and what's more interesting is that those 25% or so are most critical for synaptic functions synaptic connections so that might explain to us why the DBS finally restored the synaptic function and the synchrony of some of these key genes and proteins are really critical for that function and just to show you visually how this looks like without any stimulation the wild type and the knockout cluster separately because their gene expression patterns are different however after D-brain stimulation both of them will go up remarkably compared to the baseline but now they're intermingled so the knockout brain likewise is not easy to distinguish compared to the dentiagyrus compared to the wild type animal so these are in the null animals we were interested in the females because these are the model of Rett syndrome so we repeated the same paradigm and we collected the RNA at the same time when we did the behavior and we knew that the animals have recovered and when we see that we see that many activity-dependent genes are activated in the DBS Rett mice but most importantly you'll see that again while the females cluster differently with the sham surgery after D-brain stimulation many genes key for synaptic plasticity and neuronal function are now normally regulated at least to a similar level to the wild type and we overcome the deficits so this is at the molecular level what DBS has done so in summary stimulation of the formnix improved hippocampal learning and memory and I should mention other investigators have used phonic stimulation in rats and they showed it actually can enhance learning and memory these are healthy wild type rat and there have been studies to show it enhanced neurogenesis and there have been studies in Alzheimer to see if it can improve in clinical trials in Alzheimer as you can imagine the Alzheimer brain is a little bit tricky because the neurons are probably many of the neurons are gone so one may not get the same benefit as one in a child where the whole system is intact there is really no degeneration there is just dysfunction and I've showed you that I didn't show you the data but LTP was normalized in vivo LTP and the patterns of circuiny were normalized in neurogenesis and gene expression so at least what we can say that the rat brain at least in mice is responsive to neuromodulation whether this is going to be true for humans this is something we need to explore and we are trying to figure out ways to explore that but one thing that came out from these studies we then looked at other models of intellectual disability and this is important because today there are several hundred genes that can cause intellectual disability many of them are much rarer than rat syndrome and this year marks the 20 year anniversary for the discovery of the rat syndrome gene and we still don't have a treatment and there are many dozens and dozens of lab working on rat syndrome so imagine how hard it is when a disease is coming with a great animal model with some knowledge about the protein I cannot fathom how hard it's going to be for a disease where there may be five people affected with that disorder much more rare proteins with very little knowledge about their function so we were thinking that perhaps if D-brain stimulation can bypass the genetic defect you can go straight to circuit manipulation and that may be helpful and that's why we decided to look at data where there is gene expression data on animals that have mutations that have been seen in the human to cause intellectual disability where there is an existing animal model and data on the dentate gyrus gene expression and about quarter of the genes altered in these animals are among the genes that are highly responsive to D-brain stimulation in both wild type animals and red animals also for depression treatment with fluxity in animal models have been shown to cause increased neurogenesis and given the degree of neurogenesis we saw in our animal models we compare the gene expression data from our DBS versus fluxity and that's where we found the highest rate of overlap and I should mention in addition to the synaptic protein there were many genes that were altered that promote neurogenesis or pro-neural survival so after a new neuron is born it's actually will survive and integrate in the circuit so exercise also will do the same and last but not least people who've died from major depression and had genetic studies gene expression studies on the dentate gyrus had many of the genes or not many about 17% of the genes that are altered and responsive to D-brain stimulation altered in these people so I think it tells us that perhaps DBS has the potential and it does help in boosting the level of many genes whose are important for synaptic function and pro-survival but perhaps at the same time it can be helpful in more than one disorder and of course one has to test that what we also learn is that the brain lacking MacP2 can actually respond normally to gene expression upon the stimulation so that's one way to bypass the deficit in these animals and we hope that it promises for improving hippocampal function for other disorders with intellectual disability and to this end this is being tested now in other models of intellectual disability and there is suggestion that it can help some of them this is still ongoing work for us for red syndrome what we also like to do before we contemplate clinical trials is look at other brain regions motor dysfunction is a major problem for red syndrome so we're testing different areas of the brain to see if we can do the same paradigm but now in different areas and correct some of the motor deficits because then one can envision implanting two stimulators one for learning and memory and one perhaps for motor function and that's also ongoing work providing the benefit from DBS in the mice what we've shown that the benefit will last about three months which is really amazing that a two week treatment lasted about three months but after that you'll have to repeat the treatment so this is something we're also looking at the hardest thing is going to be how to do such trials because I think these girls have been with their disease and bypass critical period so I don't think if we just stimulate they're going to immediately learn we're going to have to combine that with some type of behavioral therapy and training and physical therapy and so on to perhaps get them to do what we want what we hope they can accomplish but this is one area we're currently focusing on until additional therapies come into play now in addition to red syndrome we learned something else about this protein we learned that doubling it can also cause a disease and this actually started in mice where as part of a control experiment we used a large insergenomic clone that contained the human gene with all its control elements and made transgenic mice so these mice had twice the normal level of the protein and to our surprise they had many problems they had autism phenotypes learning phenotypes motor phenotypes and stereotypies epilepsy and they died prematurely a third of them died by one year of life and it was having this mouse model that inspires us to look for patients who might have an extra copy of this gene and we discovered indeed people with an extra copy of these genes exist and other investigators Venesh was the first one to describe a series of those people where people had large duplications in XQ28 that had the same features that we have seen in the mice it was really the similarity was pretty striking now the human patients had large duplications they could be 500 kilobases so they spanned more than just the McP2 gene but the minimal region of overlap included McP2 and a gene Echinase IRAC1 which plays a role in immune system however I think given that our mouse only had the McP2 gene and it reproduces all the features of the disease we feel comfortable that the majority of the features of the disease are probably just driven by McP2 so in the case of McP2 I'm not going to tell you too much about the neurobiology all I'll tell you it's the opposite of the loss of function so when genes that go down in the loss will go up in the duplication and vice versa and the physiological properties are the opposite but what I'm going to share with you is the strategy we're doing to treat this disease and what Hezzi Stamburg who was at the time a postdoc in the lab and now Yanyao Xiao is completing continuing this work and the strategy is to address this question one genetic whereby we take one of the genes genetically and ask if we do this in adult mice would we correct all the features and the other one using anti-sense oligonucleotide so I'll tell you if you do this in adult animals genetically you rescue all the features so how about anti-sense oligonucleotides Ionists designed anti-sense oligonucleotides against the human gene so this way we take it out leaving just the human gene intact and this just to show you how these are wild type animals these are the duplication and here's the ones treated with the ASO you pretty much normalize the levels and you can see that by fluorescence just showing you a couple of behaviors these animals have severe anxiety they don't venture into open areas and after the ASO treatment they improve they have many seizures they pretty much we did a treatment in seven month old animals after they've had the disease now they were symptomatic for about five months and they were having continuous seizures all the time based on EG monitoring and video and you'll see with the anti-sense oligonucleotide treatment they improved so this gave us hope that this could be used as a treatment because many of you know anti-sense oligonucleotide can be used in infants and children and have been proven effective and safe in spinal muscular atrophy but we have one challenge in that here we only knocked down the human gene and we left intact the mouse gene which was a safe thing to do we went from 2X to 1X but we were guaranteed there's 1X left but in the human condition there are two human alleles so you really have to be sure you're able to titrate the dose so that you're not overshooting because if we go too low we're gonna give them red syndrome or encephalopathy and we created a new mouse model with two humanized alleles and showed that these mice have twice the level of normal and then we were able to show you can actually titrate the dose of the ASO to where if you use just the right dose you can bring it back where it's very similar to wild type of course with much higher dose you'll go too low but the point is one can titrate the anti-sense oligos just as you would titrate a blood hypertension drug for example and here is an example where you see healthy mice the duplication mice they freeze more in this contextual fear assay and then if you give them the 250 microgram that doesn't quite correct they don't correct as much as the 200 microgram than they are similar to wild type animals and this is another behavioral assay where these animals spend much longer time on this rotating rod behavior we don't quite understand why they do that but they are different from the wild type animals in this case both dosages the 250 and the 500 corrects this phenotype back to wild type level so we feel pretty comfortable that anti-sense oligos can be used and titrated to lower the level of this protein we're still working with IONIS to really ensure safety because that's still very important and trying to figure out ways where perhaps we can measure markers that will tell us how much we're driving the protein down and that's work we're spending a lot of energy on now before we move into clinical trials and through drug development so if I want to summarize one thing we've learned from all of these studies is how sensitive the brain is to make P2 levels so on this curve you'll see if one totally lacks the protein as in human males totally lacking this gene it's quite fatal of course red syndrome is right here they're mosaic 50% of the cells having the wild type allele will cause red syndrome we know from animal studies if you reduce the protein level by 50% you get a moderate phenotype and I shared with you people who have milder mutations where they have psychiatric features and this is one example of such mild mutation the LNE 140V doesn't cause any disease in females but in males it causes either juvenile onset schizophrenia or bipolar and this is the right level if you have twice the normal level of progressive neurological disease in males the females are protected because of X inactivation most of these mothers we find they have only the wild type X chromosome being expressed the cells with the wild type X being expressed and we have animal models with the triplication that are also more severe and we know humans with the triplication die much earlier in life than some of the individuals who might suffer premature death in their 20s or 30s so this is a protein that's truly graded and we've now learned from animal studies that 10% difference in the level of this protein can affect phenotype whether it's brain weight or some other features so it's really important to think about that in the context of human disease because I am sure there are mutations somewhere here where you have maybe 5-10% increase or decrease that will present with something totally milder than intellectual disability and juvenile schizophrenia as we see here and I think those remain to be discovered so having learned all this and having learned how to value protein levels from studying this protein led us to explore studies in another protein and this one is a synaptic protein SH3 and multiple anchoring repeat domain protein 3 or shank 3 for short this is a protein that's at the synapse so these are dendrites of excitatory neurons here and you'll see a dendritic spine and this protein is present in these spines and it connects proteins that regulate the actin cytoskeleton with surface protein and receptors and what we know is that it has two other paralogs shank 1 and shank 2 they're broadly expressed but sometimes there's some regions of the brain that have one and not the other so this one is abundant in the striatum and it's also expressed in the hippocampus many other investigators have studied the loss of function of this protein and we know loss of one copy can cause a syndrome of hermit syndrome characterized by some facial features as well as intellectual disabilities and autism and there are rare cases of schizophrenia and this is a common disorder accounting to about 1% of all ASD and IDD cases maybe a little bit less common than red but fairly common what we became interested in is whether just like McP2 it's also those are sensitive and whether the gain of this protein is consequential and part of the reason we were interested in that is because there were two or three reported cases of individuals that have attention deficit hyperactivity disorder restlessness unstable temper and schizophrenia and those people had very big duplications that contain many genes but shank 3 was one of those genes so those were reported two or three cases and we were interested to know among all these genes could shank 3 be the mediator and to do this we generated a mouse with an extra copy of this gene whereby again we put it under its own control and elements and asked if having 50% this is on the autism so they have two normal copies and now we've added one more if adding one extra allele will cause a phenotype and this was work done by Kihun Han and then some of the human work and the physiology was collaboration with Jimmy Holder, Christian and Wilu and this is one of those phenotypes that you actually do not need to even do more analysis to see I'm going to play the video and you can appreciate how obvious it is I think if you just stare for 30 seconds you'll realize quickly which mouse has the extra copy so it's the mouse that's constantly on the go and just will not hold still and you'll see that and these animals in their home cages then it goes on and on but of course we need to quantify that and you'll see here quantifying the activity both in males and females you'll see they're quite active we assumed immediately this is attention deficit hyperactivity like model and usually when you suspect that you have to give them amphetamines to see if that's really the correct diagnosis and when we gave them amphetamine they only got more active so their activity tripled and it was really bad so we knew it's not hyperactivity and if it's not hyperactivity then the one thing that keeps you on the go all the time is it's mania like behavior now proving to anyone that a mouse has mania that behavior is very difficult so this is where we turned back and we asked if we could possibly find humans with duplications of this gene and nothing else and that's where we collaborated with as I mentioned Christian and our genetic laboratory and discovered some patients in this particular case this one example an individual who had a duplication an extra copy of this gene that only contains this gene and one other gene but this other gene is only expressed in sperm and it's not expressed in brain so that gave us confidence that it's probably just the Shank 3 extra doses in the brain that may be responsible for his diagnosis of bipolar, mood swing, anger issues and attempted suicide one thing we noticed is the seizures we had not looked for seizures in our animals so we decided to go back and look at seizures and lo and behold we find that our duplication mice also had both EEG abnormalities compared to the wild type and seizure activities and we pinpointed increased number of excitatory dendritic spines as potentially driving that we've done a lot of studies on that what was really important is when we try to treat these animals with lithium as typically one uses in bipolar they did not respond to the lithium but they did respond to Valproate and that rescue the mania-like behavior the seizures there are many other behaviors they had besides the hyperactivity I'm not showing you all that but they were all consistent with mania-like behaviors such as being active, not sleeping as much and so on so I think the reason I share this story because knowing the underlying genetic cause of the disease can sometimes guide the choice of the therapy one of the people we identified with the duplication she was a child and children their manic sometimes they can get the diagnosis of attention deficit hyperactivity and she was put on amphetamines but she did not respond to that so that really nailing the correct diagnosis will help with the management so if we were to summarize what we've learned about the shank 3 story we know you have to have again the right amount of the protein hapluence efficiency causing intellectual disability and the syndrome and schizophrenia whereas this duplication causes mania-like behavior or bipolar and the last thing I want to share with you is this one story about one missense mutation that was found in a child in autism that we decided to study in great depth that gave us really good insight about how certain proteins function in the brain so this mutation is happened at the serine 685 it's a serine to isoleucine mutation in shank 3 this was an autism sequencing consortium that reported on this gene and they said it was in one person with autism that's all we knew about this one person so the problem is when you see that you cannot really conclude it's causing the disease because it's just one individual and it's a missense mutation we became interested in it because when we looked in vivo at the phosphorylation site in shank 3 we identified many of them and one of them was at the same serine so this would be really interesting to study and we also knew that this region is a region of interaction between shank 3 and abel interacting protein 1 so abel if you recall my schematic I said shank 3 connect the actin cytoskeleton protein and abel interacting protein 1 is one of those protein that's critical for actin nucleation and I'll get back to that so we decided this is worth studying to see if this mutation really disrupting any function and has any consequences in vivo so Lee Wang a graduate student in the lab created an actin mouse for making this substitution for the single amino acid in shank 3 and first thing he looked at is in vivo using brain tissue the interaction between abi1 and shank 3 and you can I hope appreciate here that abi1 we bring down a lot less shank 3 in the knock in animals compared to itep animals now shank 3 has many many interactors and we tested many of them and it was really only this one interaction that was affected all the other interactions we looked at were normal so now we have a protein a mutation that only affects one protein interaction and we can ask what is the phenotype of these mice and we compare that phenotype to mice that completely like the protein so mice that completely like the protein have many many abnormalities including learning and memory problems and other problems mice with this knock in allele on the other hand they had very specific phenotypes they all have to do with social behavior so if you look at these videos you'll notice that when we bring an intruder into the cage the wild type mouse leaves the intruder alone check them every now and then but leaves them alone but if you look at the bottom cage where the knock in mouse is you'll notice there's constant grooming constantly chasing the intruder and excessively grooming them and this goes on and on and on they just don't leave them alone and you can quantify that as shown here another test revealed increased social dominance if you put two mice in this tube half of the time this mouse will back up this half of the time this mouse will back up this is what happens in typically wild type animals or at least they will back up some of the times and rarely one wins in the wild type maybe 20% wins but in these animals they're constantly winning they just don't know to back up they're constantly sitting there until the other animal will back up and last but not least there was decreased vocalization in these knock in animals now we didn't only look at these things we looked at all the other phenotype in the knock out mice of Shank 3 and none of them were present in this animal so this told us this loss of this one interaction is really causing a very specific phenotype and we looked here at synaptic transmission and spine abnormality in the excitatory post-synaptic current frequency and amplitude in the striatal neurons these are the ones that only express Shank 3 and none of the other Shank paralogs and we saw some reduced and retic spine density in these neurons so now we're going to go back and ask what's happening at a protein level so I mentioned to you Shank 3 is in the spines and normally is at this post-synaptic density and it can interact with ABI 1 so what other things happen to do this we looked at the proteome the interactor of ABI 1 and we have done ourselves two types of screens to find the interactors of Shank 3 we've done yeast to hybrid as well as IP mass spec from post-synaptic density purifications from the brain tissue and we identified many actin-related protein complex proteins and actin but the one group of proteins that were shared between the two is the wave complex which is important between ABI 1 and Shank 3 which is important as I mentioned for actin nucleation so we asked if this mutant allele can interact with wave and we found it did not it could not interact with wave when you have this so this suggested to us that ABI 1 that interacts with wave is probably the reason Shank 3 also interact with wave so in the absence of ABI 1 interaction we're losing this and with this we wanted to see if disrupting the interaction between Shank 3 and ABI 1 reduced the level of the wave protein at the post-synaptic density and when we measure it in total brain extract we don't see much difference but when we measure it at the post-synaptic density where the action is that's where we see the difference in both the levels of ABI 1 as well as the level of wave because they're not being brought through this interaction so what this told us is this one interaction when disrupted can cause very specific phenotypes in human it was one case of autism but now we know in the mouse that this is really mediating the phenotype in the human and it revealed to us the mechanism by which this loss of interaction is causing probably these other abnormalities so if you look at Shank 3 it's starting to look very similar to MechP2 where you have to have the level just right hapluinsufficiency which is null allele causes syndromic autism with other features whereas this minor allele cause autism only and having a duplication causes bipolar so pulling this together I've now really learned to respect the alterations and the levels of proteins that even affects synaptic function can cause neuropsychiatric phenotype and there is a gradation based on the degree of loss of that function and we think that some rare variant might compromise just one aspect of a protein function causing a partial phenotype and we think this is likely to cause some of the more common psychiatric features so remember this one missense mutation as mild as it might seem it still caused a severe phenotype because this was a child that presented with autism so this person presented early on in life with features when you think about a psychiatric disorder such as schizophrenia where the person may go through childhood very well and only present when they're 20 years of age it's going to probably be even a milder mutation and I think this is the challenge facing psychiatry today how we're going to find those genes where a very small subtle perturbation may be causing disease and I hope I convinced you today that rigorous evaluation of good mouse models can inform clinical medicine so in the case of micro 2 duplication it predicted the human disorders in the case of shank 3 duplication it showed us the mania that then allows to discover the patients and this missense mutation corroborated the pathogenicity of a mutation and I think what's really important is going back and forth between the human and animal models we're always inspired by the patients but one has to study them in animal models and look for phenotypes that are true to the disorder that are meaningful and eventually as in the case with the people with bipolar and seizures it was the patients that let us to go back and look for seizures in our animal models so this back and forth is really important and I find that it's the only way really if we're going to make a difference for either discovering mechanism of pathogenesis or eventually come up with some precision medicine and as I showed you in the case of shank really overdosage there are some situations where you can actually change the management based on understanding the genetics so with that I'd like to thank all the people who've contributed to the work I always like to thank Ruthie the postdoc in my lab who discovered the red syndrome gene after three years of sequencing gene after gene to no avail she finally found it and then I mentioned all the other contributors to the work our collaborators and special thanks to the families red syndrome might be two disorders and shank duplication families and of course our funders thank you very much so let me just say that I'm sure that who will be taking questions but afterwards there's going to be a reception over in biochemistry so you're all welcome to join us to talk to her in more depth over there so we'll take some questions now after yes sorry I skipped details to cover more grounds the way we delivered it the first time we delivered it in the published paper using pumps where we put alzat pumps oh the question sorry how did we deliver the ASOs so in the published paper we used pumps so it was delivered over a period of six weeks but in the new work we just used intraventricular injections and this sort of mimics what would be done in a human down the line which is a spinal tap intrafecal introduction and what's surprising again I didn't show all the data here but it does have a broad distribution so it reaches different regions of the brain and the degree of reduction of my P2 is similar throughout the brain maybe the cerebellum a little bit less but other than that it's pretty similar specific and only one are there other similar syndromes with other chromatin regulatory many yes so is this just one example and do they cause autism and schizophrenia as well yes well schizophrenia we don't have as many genes for schizophrenia but there are many genes that affect chromatin from enzymes that affect histone modifications to other chromatin remodellers that have been discovered as causes of autism in fact it's among the more common categories of genes that cause autism and intellectual disabilities so the question is whether brain unique expression is what's driving the phenotype and the answer is no these are broadly expressed proteins my P2 is expressed in the peripheral nervous system is expressed in the peripheral tissues and many other tissues but we do know that the phenotype comes from its loss in the nervous system because if you only do a knockout I mean I shared with you the inhibitory and excitatory neurons that's enough to reproduce most of the features of the disease but even if you do it in the whole brain you get similar features so although they're expressed in the peripheral tissues the phenotypes are brain phenotypes so the question is are there changes in the levels of mitochondria there we have not been able to see much changes in the mitochondrial function and or if you we didn't really measure levels but at least from a functional point of view there haven't been many changes in the animal models that we've been able to show in the past people have looked in humans and for some times people suspect that there could be mitochondria dysfunction the data are mixed there may be some changes but there's so much disuse and decreased activity it's hard to tell what's really primary and secondary I don't think of MECV2 as a mitochondrial disorder because most people have not been able to pinpoint any mitochondrial specific deficits so the question is is there any brain stimulation at least in the hippocampus in the phoenix to help the hippocampus would that help any disorder where it could be synaptic dysfunction and this is what's being tested right now at our center Jean Rang has looked at CDK-L5 disorder where they have learning and memory and he's also looking at additional disorders that challenge is many of these human diseases are not as well modeled in the mice so when you go back to test for behavior we can't either replicate what's been published or we don't see much of a phenotype but I think that data are hinting that it can help other disorders and until he tests this in two or three models we will know so it's possible that if you have decreased synaptic plasticity and some decreased function in learning and memory that's associated with decreased synaptic plasticity we do know that this treatment is enhancing synaptic plasticity and neurogenesis so a disorder that may benefit from those will probably benefit and the examples I showed will be probably some diseases that could benefit because we know that genes that are altered in these examples are corrected with DBS so what we know anatomically it makes more new neurons we have not really measured synapses we just measured the physiology the plasticity and we saw that that's normalized but we have not looked at synapses per se in the animal models and this is where we're trying to find the optimal way to repeat that because sometimes if you do this and then you come back to the same animals it's hard to do some of the same testing on them you sometimes lose the actual phenotype so this is what we're optimizing right now to do that I'd like to know if you can do it for two years at least through the lifespan of the mouse and what happens so this is really the big challenge and conundrum at the molecular level and part of the reason I didn't dwell over our molecular study because I still think it's work in progress it is surprising that there are more genes that are down-regulated than up-regulated and what you see let's say they're about we find that in some tissues we see about 1500 genes that are down-regulated the same genes will be up-regulated in the gain of function model the duplication model so the question is why is that how much of that primary how much is that secondary and as much as we've tried to really to get at that point it's been very hard so we're going now to many additional biochemical studies to find more interactors what could be modifying this protein and to really do some time studies to really identify the earliest possible changes to really explore what's primary and secondary but you're absolutely right if you just look at the gene expression patterns it does not look like a classic and we don't quite know why is that so it could be that in different contexts it does different things it represses very few genes and all the other things are secondary it could be that it represses some and activates some and we're exploring all these possibilities one more time