SYNGAP1 Epilepsy Research: Insights into mechanisms and therapies

Here are the introductory comments: 

In this video Dr Shilpa Kadam presents SYNGAP1 Epilepsy Research: Insights into Mechanisms and Therapies. Originally shared on July 2nd, 2020. Please find additional Syngap research webinars here: https://syngapresearchfund.org/webinars​

Dr Shilpa Kadam is an Associate Professor of Neurology with joint appointments at The Kennedy Krieger Institute and Johns Hopkins University School of Medicine.

Dr Kadam’s lab is focused on investigating mechanisms underlying epileptogenesis in acquired and genetic developmental disorders like SYNGAP1. The lab is interested in discovering novel drug targets and delivery systems directed at the central nervous system.

Dr. Kadam has broad clinical exposure to epilepsy management.

Her work with SYNGAP has been focused on the relationship of poor-quality sleep and epilepsy. Her laboratory was able to identify abnormal underlying brain activity and high incidence of seizure during sleep by examining mouse models and overnight EEG from SynGAP patients. Her laboratory then discovered that using a low dose of Perampanel (Fycompa) may not prevent seizures but can help correct the high-frequency abnormal brain activities during sleep in mice model. This itself may have beneficial effects during sleep of SYNGAP patients. She hopes this approach will lead to a human trial.

BiologicalPpsychiatryJournal.com/article/S0006-3223(20)30002-0/fulltext

Dr. Kadam is also interested in antisense oligonucleotide therapy ASO to help rescue SynGAP levels in the brain.

She is currently seeking funding for several projects in SYNGAP with one of the ones more important being on CRISPR/Cas technologies to create mice models to find potential biomarkers.

‍

Transcript

Dr Kadam: My training  has been in epilepsy research and it was Dr. Huganir who introduced us to SYNGAP1  research and the mouse that the data we will  I'll be showing today and we will be discussing is  the is the mouse that his lab generated quite a few  years ago but it has a very critical mutation  in a very critical domain that allows us to  study almost the most critical functions for SYNGAP1  which has been very well documented by all  his early research and continuing research. So  what I'm hoping to try and communicate today is  how we approach this research. Why did we do it the  way we did it and some of the insights that are  maybe very novel that might generate new  hypotheses that my lab, myself will pursue in the  goal of therapies. And as parents you guys know  that, you know, understanding how this protein is  affecting the brain development is very critical  but at the same time if we can try and figure  out what new therapies might help improve, even by  a little bit, the incidence of seizures in children  with SYNGAP1 mutations is also important.So my slides are not moving forward... okay. So  all of this information is known to you but  I'm going to give you a background just so that  you understand what we are focusing on. So just  like Sydney said seizures and sleep disorders are  predominantly one of the spectrum of SYNGAP1  haploinsufficiency and myoclonic and absence  seizures are the dominant phenotype both described  in Ingrid Scheffer's first cohort study and I see  it as the description for all the reports that I  get for the EEGs that the parents have sent  us copies for so far. And of course in epilepsy  research there's a long-standing relationship  between epilepsy and sleep problems especially  in neuro development disorders.

So at Kennedy  Krieger we have huge focus on Rett Syndrome and  and there are many similarities in the sense of  the progression of the epilepsy, the progression  of the learning disabilities, and some of the  sleep disorders that have been reported by  parents and have been phenotype by the clinicians  who are looking at cohorts of patients with SYNGAP1  mutations. Sleep disturbance. I don't have  to report it. In 100% of the patients problems  are associated with sleep initiation and frequent  night awakenings. Early morning seizures has also  been reported and so this will become important  as we try and describe some of the results we  found from the analysis of overnight EEGs that  parents have sent us and what we then found in  the mouse model. Because for very obvious  reasons once we have this phenotype there are  certain investigations we can only do in the  mice. We cannot do in the children. So that's where  translational research becomes so important. So  just as a quick background I know many of you have  taken your child for clinical EEGs and overnight  EEGs so you may have seen some of the trace recordings  either given to you in the report or you  may have actually seen with the EEG recording  and I just wanted to give you a little bit of a  background and hopefully you can see my mouse.

So  basically right now we are focusing on wake versus sleep EEG and so wakefulness, the circuits that  in our brain that promote wakefulness, usually have  these very low voltage, fast oscillation kind of EEG  phenotype which is right here. Wakefulness you see  there is very low amplitude, high frequency  discharges on the EEG. And then there are the three  stages of sleep in humans, which is stage one, two  and three and you can easily see that as the non-REM sleep, which is non-Rapid Eye Movement sleep,  it becomes more dense, you start moving from this  low amplitude, high frequency in stage one to this  very high amplitude, low frequency type of sleep  brainwaves. So these are very characteristic of  non-REMs and then we have REM which is Rapid Eye  and so the brain alternates between non-REM and  REM during our sleep cycles in humans and REM you  can see is almost similar to wake. So it's called  paradoxical wake, paradoxical sleep  because the waveform if you just looked at it  and you don't know that the child at that  point is awake or asleep on video you might confuse  it to being wakefulness but the significant  distinction is that a REM will always be  between two non-REMs. So when an non-REM ends or  REM starts and then a non-REM again progresses  from stage 1 to 3 of non-REM and because these  are associated with frequency oscillations on  how the brain is firing in certain frequencies,  I put in this slide on the side in blue which  is all a one-second discharge and you can see  what we call as "high frequency oscillation" which  is gamma, which would be the focus of most of the  cortical function we are trying to evaluate, both  in the mouse and in the human EEGs. And so as you  can see, this is high frequency. It's around 30  to 100 Hertz. Then there are other oscillations  very well defined in the clinic and we try and  apply the same rules to the mouse EEG. And so as  you go lower in frequency it's the beta, alpha,  theta but more importantly, Delta, which is the most  predominant, just zero to four Hertz, is the most  predominant waveform seen in non-REM stage four.  So deep sleep is associated with a high density  of delta. High brain activity associated with doing  some tasks, being cognitively aroused is associated  with high gamma.

So for this talk we are going to  focus on gamma versus delta because they are the  predominant frequencies in wake versus sleep.  Okay. So then we'll just start straightaway with  what we found with one of the overnight EEG's  from this is a three-year-old child with SYNGAP1  mutation and I'm going to focus here on the  middle panel where we have done the power analysis  for the frequency, we had the video also, we had the  notes from the clinic. And so if you just focus on  so here on the x-axis this is a 20 hour long EEG,  including the night. So let's say it starts late  in the morning and goes on to early in the morning  the next day. Okay. So that's the x axis. On the y axis  are the different stages that have been scored  for this EEG. So here's wake. Here's non-REM. Here's REM.  So this line over here as this continues is when  the child is awake. It jumps up to non-REM when  the child first falls asleep and goes into the  first non-REM cycle. Then as we go forward over  the entire time, which is quite a few hours that  this child is asleep, you can see that the child's  brain is switching between non-REM, REM, back to  non-REM and it keeps doing that between multiple  cycles until the child is awake in the morning  and that's the wake state again. So basically this  entire jump here is the time based on the EEG and  the video and announced the child was asleep. So  then overriding that all the raster plot of  the red lines shows when this child had episodes  of short duration 3 Hertz seizures. Here is  one example from... let's see, from this one event  of this high amplitude 3 Hertz spike wave discharge  on every channel. So this is a global discharge and  from the background EEG which is noisy which would  always be noisy when the child is awake because  they're not going to stay still but what you can  see in the background is this very low amplitude/  high frequency discharge which is indication that  this child is awake. The cortex is in mostly in  gamma frequency but in the middle of this gamma  frequency this child has 3 Hertz high amplitude,  short duration, 2 or 3 seconds. So that's an example  of that event happening when the child is awake.  Then here there's an expansion of a similar event  occurring during sleep. Just on the background you  can make out wake versus sleep because you have  these high amplitude delta waves in the background  but the seizure episode is exactly identical.  It's 3 Hertz high amplitude, short duration.

So  that is the predominant phenotype of  seizures on this EEG and when we identify every  episode that happened in the 24 hours you get  these that there's an episode and as the child  is coming closer to the time, even so at this  point the technician in the room, the parent is  does not know when the child is going to fall  asleep but the brain knows when the child is, when  the brain is moving towards sleep. And then you can  see that there's an increase in the frequency of  these discharges and then there's a short burst  right when this child's brain transitions from  wake to non-REM and that's here. So these are  multiple episodes. That's why that line looks  thick. Then what happens as we go through and as  we scored them independently and just superimpose  them, right when this child's brain, because now the  child is asleep so there is no outside influence  that would trigger a seizure, but inside brainwave  states the child's brain is in non-REM, it  switched to REM, that's when the brain throws a  seizure. Then the brain is in REM and is switching  to non-REM exactly at that transition point the  brain throws a seizure. Again, another transition  point there's a burst of seizures. And then again at this  transition point there's a burst of seizures. And then when the  brain knows it's going, the cortical state is  moving from sleep to awake even before the child  actually wakes up, the brain senses the strongest  transition state and throws the highest bunch  of seizures in this EEG.

So the point I'm trying to  make is: most of these seizures are ramping up when  the brain is transitioning either from wake to  sleep or from sleep to wake and, very interestingly,  even between sleep stages where there would not be  any outside influence of light or movement or TV flash  or some music. When it's transitioning from non-rem  to REM and REM to non-REM this brain throws this  3 Hertz discharge. And here I have this lowest  panel is just an expansion of this, what we're  going to call a cluster, that now as the brain, as  the cortex realizes that this brain is going to go  into a wake state, it throws a multitude of these  shorter shots they look just looks because it's a  tighter timescale but I'm just trying to show you  where four seizures within this cluster (they are  not even all the seizures happening that cluster)  start happening repeatedly just before the child  is awake. So these might happen just when the  child is going to wake up or it might happen as  soon as the child is awake which is I think what  the previous reports from the clinicians are where  there is reporting from the parents that there is  early morning seizures, difficulty falling asleep.  

Now as the brain is transitioning, if the brain is  seizing, there's lots of literature to figure out  why then that child would not transition easily  and might have difficulties falling asleep. But  these are hypotheses at this point and  this kind of interesting distribution is what we  are trying to figure out. If we can understand  better and then when we test novel therapies  we can, in doing the same analysis, figure  out how that is changing or not changing. Okay. So  obviously the translational version of it we  use a mouse model of SYNGAP1 and this mouse  model like I said was generated in  Rick Huganir's lab. What we are doing and what we are  good at is doing these 24-hour continuous video  EEGs with your EMG and so we have a 12 hour light  and dark cycle, so making both the daytime and  the nighttime. The technical portion is how we -  this EEG signal - how we convert it here with FFT  so that we can analyze all the frequencies within  it. And then of course what was the title of a  recently published paper is this FDA-approved  drug that we tested, perampanel, which is the  AMPA receptor antagonist. And we chose this  drug because of all the literature that Rick's  lab has already published showing that SYNGAP1  at the postsynaptic membrane, having this huge  impact of how many AMPA receptors are inserted. Now  AMPA receptors are, they receive the excitatory  signal, so if there are too little or too many  very simplistically it would completely affect  how that circuit is functioning. So just not even  as a therapy, we wanted to test. This is the  only drug available that has this specific  mechanism. And since it was approved for epilepsy  very recently we thought this was the best drug,  given the data from Rick's lab, to test in this  mouse to see if we blocked AMPA receptors what  effect it would have on the EEG and especially  on the seizures.

So first what I'm showing you is  video of one such seizure that we recorded in  a SYNGAP1 mouse and hopefully I'll be able  to play it again but what this is is  the early EEG was during sleep. The mouse threw  a seizure and then it transitions to - it has some  spikes - but it transitions to awake EEG. So I just  showed you the clinical EEG and how we scored it  and about how these seizures were happening  during transitions between sleep and wake and wake and  sleep and so here's an example of a phenotype.  Maybe I can play it again. So you can see where  the seizure happens and these are very  equidistant spikes associated with spikes on the  EMG showing that it's an myoclonic seizure. It's a  short duration seizure and the frequencies  very surprisingly was 3 Hertz. So this is a 3  Hertz seizure happening at a juncture where the  mouse brain is transitioning from sleep to wake.  

So now the "chicken and egg question" is: did this  mouse brain throw seizure and that's why it woke up? Or did the mouse brain throw seizure because  the brain was transitioning and whatever is  the mechanism by which the cortex changes its  sleep versus wake state that makes that brain  susceptible to throw seizures? So those are  the questions that came to our mind when we saw  very consistently the Syngap mice throwing  seizures at these transition points. And then  we were excited with the phenotype because it's  it's a 3 Hertz seizure, it's a myoclonic seizure,  and that is the predominant phenotype in the  patients. So here is the same seizure without  the movement so you can focus on the leads. So the  first lead is the EEG lead. What I'm trying to show  you are the spikes that you saw when the mouse  was seizing. On the bottom is the EMG and this  is to show you that when the cortex is spiking,  throwing a spike, the muscles on the back of this  mouse, on the neck and shoulders of this mouse,  is throwing a spike (with a slight delay). So that  would be the transmission of the cortical neurons  firing and the muscles in the back contracting. So  this had a nice correlation and  the time scale here shows that this is a 3 Hertz  seizure. On the bottom is just another example of  a seizure where there are these random spike discharges  but you can see the EMG is completely silent.

So what does this mean? That there are seizures that  you will identify in your child because there will  be a motor component either the eyelid myoclonia that  Ingrid Sheffer has described in their  paper. So there's a motor component and if  you are near your child and you're observing you  will know that something is wrong and  that there's a seizure episode. What this data  on the bottom is trying to show you is that  there can be seizure activity in the cortex of the  mouse for sure and the EMG silent. Meaning there is  no motor component and so these could be  what we call "electrographic only" seizures so  they would be silent seizures. So there may be a  component where there is seizure activity where  there will be no behavioral component and  those are silent seizures which is troublesome  because if there's a huge number of these the  seizure burden on the cortex makes that cortex  malfunction even more because severity of seizures  can affect how the cortex or the brain functions.  

So this is the quantification of the seizures we  found in the entire study and so it's basically  because we were doing it in two-month-old mice,  which is the P60 and then the same mice when they  were older (P120) and you can see that over time  when we recorded the number of seizure events  went up and the circle here denotes the  myoclonic seizures. So the younger mice had the  specific just single phenotype but as they got  older they started showing multiple different  phenotypes. Okay. So this is what we would call  progression in the mouse model. And this  pie chart here is showing you that most of the  seizures were starting in non-REM. And this is very  very well known in both in epilepsy patients and  in development disorders associated with seizures  that there's something about the non-REM state  with the slow oscillation Delta wave that makes  the brain most susceptible to throw a seizure. So  that was the phenotype in the young mouse where  most of the seizures were arising from sleep and  then as the mouse got older you can see that that  frequency changed. Most of them were then happening  during the wake state. So all this is saying is  for the model that the phenotype is changing over  time. And we know this to be true in developmental  disorders that the epilepsy is not static. It  sometimes progresses, sometimes it regresses and  then sometimes the type of seizures change. So  this is interesting for the mouse model  because then now we are having multiple seizure  types and if we do interventions we can see the  effect of multiple interventions on those various  phenotypes. So this is all good for the modeling  world.

Mike: Dr. Kadam, I'm sorry to interrupt you  but if we could just go back to slide eight for  a second. For those of us who don't  spend a lot of time thinking about mouse models  would you mind just elaborating on what P60 and  P120 mean and how those how those ages translate  to kid ages because this is clearly...

Dr Kadam: That's a very  important question. And I was definitely going to  touch on that because Brennan is on this call and he's working on, and we are working  on, doing much younger ages because of course  this is a developmental disorder. So this was  a pilot study. A P60 means 60 days old. So that's a  two month old mouse. So that's a young mouse. So  mice are weaned from their moms or "dams" at around  P21 right. So that would be the time when when they  can fend for themselves. So P60 is a young mouse.   It is definitely not a child though. And P120 is  just a older mouse and here the significance is  because then we're trying to record from the same  mouse to look at progression. So as a pilot study  and coming from the epilepsy research field there  is a very important caveat where even if you  look at other papers where people have studied  Rett syndrome, CDKL5, you will see that they're  using adult mice. And more and more as we have  these grants for review and reviewing grants,  NIH and the reviewers are like "these disorders are  developmental so you need to start looking at much  younger ages". Luckily that is the bread and butter  of our lab. We study neonatal seizures  in pups but because this was our first  initial foray into SYNGAP1 and nobody  had quantified how the epilepsy... what was all  the phenotypes possible in this model and how it  was progressing over age. We needed  this infrastructure to start then going uniquely.  Because neonatal EEG is tricky and I will talk about  that as because those are the experiments we are  doing right now. So hopefully that has addressed the  question but I will touch on this again when we  start talking about what we are doing currently.  

So that the ages we are doing right now will be  covering the age even from a newborn to the first  two years of life where we will want to figure out  what is happening in the same mouse at early ages  and can we do interventions that would then change  the trajectory? But now we know what the trajectory  is. So if we did some intervention in the pups  and then we did a P60 again and lo and behold  they had no seizures, then we we could convince  the NIH or pharma that look, our intervention is  changing the progression of this disease. At least  for the epilepsy. And then when we talk about the  gamma oscillations which totally underlie  cognition that we are actually trying, we are  improving cognition in the mouse.

So this data is  critically important because in translational  research we need this end point to  show data that would have NIH's confidence or a  small big pharma's confidence that we are actually  by some intervention, novel or, like perampanel,  that's already approved, we are able to change  the trajectory of the disease. Because developmental  disorders have that, that in an immature brain  having seizures there's lots of  literature to show the seizures themselves affect  how that brain matures. And so it's difficult to  parse out how much are the seizures by themselves  contributing to the cognitive dysfunction versus  the underlying genetic cause that may be  contributing to cognitive dysfunction.  

So "part and parcel" of seizures are interictal  spikes. Interictal means between two seizures. So  any kind of spike activity that is now identified  by these red asterisks here is called interictal  because now we know that this mouse  seizes and so interictal spikes by themselves  are a huge area of focus and research for being  able to predict when a seizure is going to happen.  So usually normal EEGs will not have these spikes.  Or at least not at this frequency. So what this  very simplistically indicates is that this brain is  hyper excitable. There are times when the person  or the mouse is just normally functioning that  a bunch of neurons just throw a massive spike  together. So this does not happen in a normal brain.  This is just an example of spikes happening during  wake and spikes happening during non-REM which is  the sleep state we just discussed. Here again is  the background EEG showing you wake. It's small  amplitude higher frequency. Here's the background  showing you high amplitude, slower frequency.  The spikes almost look identical but what  this example shows you that in general we found  there were more spikes during non-REM then during  make states. This again ties and is very common in  epileptic brains where seizures and spikes happen  during sleep states and that that circuit state  is more conducive for seizures to appear.

So here  Brennan went and quantified every spike in the  24-hour EEG. So this is a lot of work and it shows:  so the wild-type is the mouse that does not  have the mutation and then the P60 and P120  like we discussed is the mutant mouse at 2 months  old and then the same mouse older. And you can see  a natural progression in the incidence of these  spikes but you can see it as the color code is  wake versus sleep that most of these spikes and  significantly are occurring during sleep. Then  we also know that, you know, the light phase  has an effect on spike and sleep generation.  So because mice have diurnal cycle in addition to the  circadian cycle, when they sleep during when the  lights are on versus when the lights are off. Why  is this important? Because unlike us rodents are  nocturnal. So basically they're sleeping during the  light and they're exploring, eating, grooming during  the dark. And as you can see, most of these spikes  when we parse the same data into whether it was  during the light sleep cycle or the dark sleep  cycle, most of them are occurring during light  sleep. So this is when the rodents are  sleeping. So this would mean like a child during  the night if we were trying to make a  human comparison. And then as we have already seen  before when you look at the sleep stage within  light sleep or dark sleep, spikes are driven by  non-REM, without a doubt. This is not very unique  because Rett syndrome kids will show  the same thing. Most developmental disorders that  have epilepsy this matches the profile that non-REM  states make the brain more vulnerable for  throwing spikes and seizures. And over here so  this bar on the top is showing you the start of  the EEG and end of the EEG but the light cycle  on. So here the lights are off. Lights are on.  Lights are off. So this the black portion adds  up to 12 hours. The white portion adds up to 12  hours. And so here's the 0 to 24 hour indicating  the duration of the EEG. This is the frequency plot  for the number of spikes and you can see how they  go up and down at P60.

So this is a younger mouse  and then the same mice a little bit older. So in  general you can see the frequency of spikes are  going up but when the seizures happen, which are  these red lines, when the seizures happened  in the younger mouse there does not seem to be a  direct correlation between when the spikes were  happening and when the seizures because there  seems to be a continuous occurrence of spikes. But  then when you look at P120 you can clearly  see what we have seen here with a non-REM and in  the light cycle, there's a little bump during the  light cycle which is this white bar here. So  frequency of spikes is going up during their  sleep during the light cycle and the seizures start  occurring when this transition is happening. So as  the spikes are going up the seizures are going up.  So what all that could mean at this point is that  susceptibility for spikes and seizures starts  going up at that time point in the light sleep  during non-REM. And this is just interesting to  keep tabs on because as we start analyzing more  of the EEGs coming from the kids with Syngap  we will want to see whether this holds  true for the kids too. And this has tremendous  relevance for when an anti-seizure drug is given  to the child. So let's say, just a very  simple example, if this is a 24-hour cycle and  obviously you're not going to wake up your child  and give them medication. Well that's not usually  the protocol. But if the seizures are happening at  that time and you give your medication here versus  here it's very easy to hypothesize that this dose  given here maybe would be the early morning for  the child will not be efficacious versus if you  gave it here maybe it would stop all of these  seizures. So I just want you to keep that in mind  on if the seizures are occurring at a particular  time of day but the dose and then the highest  concentration of the dose in the body is at a  different time of day that would for the same dose  have completely different efficacies in the same  child.

Okay so this is a hypnogram from a mouse. So this is what I mean by the...  so here's, this is 24 hours, this is the light  cycle. Here's dark and then the light. So daytime.  Nighttime. Daytime. 12 hours each. And the same thing  as we showed before. So the yellow is all the  time that this mouse is awake. The blue is when  the mouse is in non-REM and the red is when the  mouse is in REM. So that's wake, non-REM, REM. So you  can clearly see obviously there are differences  between a mouse hypnogram and a human  hypnogram but the stages are similar. The transitions  are similar and when I said ultradian is that  you see that because this is a light cycle that's  when the mouse is sleeping and hiding away from  predators there are many more sleep cycles. During  the dark cycle when they are very active exploring,  feeding, grooming, they're awake for most of the time  but there are sleep cycles during  the dark phase and then it moves again to the  light phase. So basically you are now familiar with  the three stages: wake, non-Rapid Eye Movement (non-  REM) and then REM. I already said this, the mice are  nocturnal and so they have the ultradian each  sleep cycle. So hopefully now you understand what  the difference is between a mouse hypnogram and a  human hypnogram. So when we looked at these mice  and their sleep cycles because sleep is a huge  problem in many development disorders including  SYNGAP1 here is a wild-type hypnogram. Here's  the one of the younger the two-month-old Syngap  mouse and then the same mice when they got older.  And just eyeballing it if you see the same mouse  at P60 versus P120 you can see the difference  in how this mouse is now cycling. And when we  quantified that, because mice are nocturnal and  that is critical to their survival, if they start  being active during the day and sleeping  most of the time when they're supposed to be  exploring and feeding you can see how that could  be detrimental but when you quantify that you can  see the wild-type there's a clear distinction of how  much percent of the time they're in wake or sleep.  At P60 this difference goes away a little bit  but it's still significant.

At P120 this mouse  has completely lost its nocturnal predominance  of how it does its activities and that the same  version, this is the same data quantified in a  different way. So basically you can see for light  and dark, red versus the blue, the wild-type, as  you would expect a normal mouse, has a significant  difference on how it cycles through in the light  and dark cycle. And the P60, P120 this  is lost to a greater extent. So this of course  we have to see in the light of a mouse how this  would be detrimental for the mouse's survival and  well-being. So the take-home message from here is  that the sleep cycles are completely disrupted  in the Syngap mouse compared to their wild-type  counterparts. And then here we have, this  trace, you know, when the mouse is in the cage the  video is on there's software where you can trace  all the movements the mouse does  during the 24-hour cycle. So this is qualification  of once a wild-type and a Syngap mouse. Here's wild-type. Here's P120. And the software keeps  tracing how the mouse moves. So you can see that  during the day, during the 24-hour cycle the  wild-type has this amount of movement. You can see  a hotspot here which I would predict would either  be where the waterspout is or where the food was.  But look at the same so both this wild-type that is at  P120 and look at the Syngap mouse. This tremendous  hyperactivity. The hotspot is in the center here  which would almost mean it's not associated with  with where the food was or where the waterspout  was. So it's just overall hyperactivity. And when  you quantify that based on the light cycle where  here's the the period when it's dark and this is  normal so the if you follow the black line which  is a control which the normal mouse does and  so they are very little active during the day  and when they are nocturnal their activity goes  up and this is when they do all their movements,  all their nesting, grooming stuff and then when  the light comes on it goes down again. And then  in the dotted line is the younger Syngap mouse  and the solid red red line is the older Syngap mouse.  And you can see there's some hyperactivity even during the cycle but during the night this activity is very  significantly high and almost when the lights come  on at the older ages they don't even stop. They  keep going before it comes down.

So the take-home  message here is that these mice replicate some of  the phenotypes in the Syngap Syndrome where  there's tremendous hyperactivity. Now I mean we  will be doing additional tests to see whether this  activity is directed or just not related  to any task and here is the quantification for  the things showing you the significant increase  in hyperactivity in the Syngap mouse. Both at P60 there's  a trend and then becomes very hyperactive at P120.  So then we move. Now we transition. So now we  know this mouse has seizures. We are excited that  it has a phenotype of seizures that is actually  seen in the patients. We are excited to know that  they're happening at the transition points that  our early human EEG studies are showing that  they're occurring at transitions between wake and  sleep and non-REM and REM. So like I said there's  a huge area research of great researchers who  are looking at gamma oscillations in the cortex  because it underlies cognitive ability. If you are  focused on a task your gamma oscillations in the  cortex go up. Even if you're awake and you are not,  you know, you're in a resting stage you're gamma  oscillations will go down. So there's a very tight  correlation of task engagement, cognition, learning  and gamma oscillations in the cortex. Like I  showed you before it's a 30 to 100 Hertz in  mice because using protocols from our  previous research we are looking at this 30 to 100  Hertz kind of gamma rhythm and then there's lots  of literature to show that gamma oscillations are  driven by interneurons. Now inter neurons are  the neurons that are the "no-go" signal in  the brain. So there's excitatory neurons which are  the "go" signal and the interneurons that are the  "no-go" signal meaning you stop the neurons from  firing. And one class of these neurons is known  as parvalbumin (PV). You will see this  PV in all the next slides but it stands for parvalbumin interneurons and there's lots of  literature to show that these interneurons they  drive gamma oscillations in the cortex. So here's  a simple schematic and I'll go through this again  later on.

So what I just said that there are  excitatory neurons meaning the ones that fire  signals for us to do an activity if it's motor or  to start a start a rhythm for certain functions  in the brain and those are represented by these  blue triangles. So the PYR is pyramidal neurons  these are excitatory and they are in the cortex  in huge numbers. Parvalbumin neurons it says GABA  because that's the ligand that's released by these inter-  neurons onto the excitatory neurons that gives them  the "no-go" signal. So when this neuron is firing, if a parvalbumin positive neuron fires onto  it (and that's why it's a "no-go") it will stop  this excitatory neuron from firing. So that's  very simple. Excitatory: go. Inhibitory: no-go. The red  ones are the interneurons the blue ones are the  excitatory neurons. But to make this slightly more  complex, these interneurons are connected to each  other with electrical synapses meaning they can  then now fire together much faster instead of them  communicating to one another but they also inhibit  one another. So this is a very interesting  circuit and this subgroup of interneurons  is very critical in the cortex and we will go over  this a little bit later. So for now right now  what we're looking at is the "go" signal neurons,  the "no-go" red neurons, they inhibit the excitatory  neurons and they talk to each other. They inhibit  each other and they also talk to one another. So  that is the circuit in the cortex that many people  have focused on because they underlie these gamma  oscillations in the brain. So these PV interneurons are also called as fast-spiking interneurons  and I pull this slide just so that people  who have not seen firing patterns of these "no-go"  signals can understand why this is a special class  of interneuron. So here is patch clamp onto an  interneuron and this B graph is showing you  the single firing pattern of fast-spiking  interneuron, which is our PV interneuron, a non-fast  spiking interneuron, which are the other types  of interneurons and a pyramidal neuron which was  the blue neuron in a previous slide. So here's  a fast-spiking. So this, if you take anything  from this slide, is PV interneurons have a very  high frequency firing way. What that means for  electrical properties and how this neuron is built  and is all interesting research that's ongoing  and then you can see a non fast-spiking  when it goes fast and then it peters out. And your  neuron is a much slower firing pattern so very  simplistically it would be if this blue neuron,  which is an interneuron, has to stop this neuron from  firing you would expect that it would have to fire  it a much faster rate then the excitatory neuron.  So this neuron would fire at this rate and stop  this neuron from firing. And then the other (if  I can go back a slide) the other characteristic of  the specific parvalbumin interneurons is that  they inhibit the neurons right on the cell body.  It's like before the horse gets out of the stable  it's a "no-go" signal. So it's a circuit  which controls very tightly the firing pattern of  the excitatory neurons. Other interneurons are  known to form synapses on the branches of the parvalbumin neurons.  So this is a very specific characteristic of a fast- spiking interneuron. And then we just talked about  the firing pattern. So this is just to show you  that these interneurons have the ability  to spike at very fast frequencies compared  to the excitatory neuron and other interneurons.  So now this is the same data that we saw for the  hypnogram before and we have color-coded it for  the gamma power on the EEG for the mice. Okay. So  the color code is still the same. We have taken  out REM because in the mice they're very short  duration. So basically this with the same color  code this yellow color is the wake. Blue is the  non-REM but instead of seeing the hypnogram  now you're seeing every dot here is the 10-second  EEG power for gamma frequency. Just for the gamma  frequency.

Okay. So here's again the  bar on top for the 24 hours. This is switched  around for the way in which the data is  shown. So it starts in the night. Here's the entire  12 hours of light and then here's the remaining 6  hours of dark. So what you can see in the wild-type,  you can see this mouse is going through wake  cycle, sleep cycles, wake cycles, sleep  cycles but what is happening to the gamma power  when the mouse is waking up and sleeping? As  soon as the mouse wakes up the gamma power goes  up. As soon as the mouse sleeps the gamma power  starts falling. Okay. So this makes sense for  the simplistic thing I said that when you are  awake and active and doing something or reading  something the gamma oscillations in your cortex  will go up. When you sleep and as the delta  slow waves start coming in, gamma power comes down.  So this is a way how the brain adjusts and it's  known that this is very critical for  learning. So whenever the gamma power goes up and  your're awake and you are learning a new thing, you're  reading something new, it is very critical that  your gamma power drops during sleep so that the  brain can then put that what you've learnt or what  you've read into your long-term memory. So that you  remember it the next morning. Or with multiple  readings in a few days. So that's how the brain  learns or this is one of the components that's  been proposed underlying learning and memory. So  consistently, and this is just an expanded version  of the data in this red box, here the mouse  is awake. As soon as he falls asleep his gamma  starts dropping. Even before the mouse is going  to wake up, just as the child will wake up, the  gamma starts going up. Okay. And you see that  consistently. So you see the gamma oscillating up  and down, up and down, during sleep. What happens in  the Syngap mouse? So here is an example of the same  analysis from the EEG of a Syngap mouse and you  can see it's color-coded for sleep states but we  don't- the oscillations become very muted and in  fact if you watch closely here and here... here is  the gamma during wake and it's actually jumping  up during sleep. So it's the relationship of what  gamma should be doing in the cortex is not only gone  away but in certain cycles it's actually switching.  So here you can see here's the gamma. It goes up  during when this mouse falls asleep and actually  goes down when it's awake and then in one cycle  it actually goes up and so the relationship of  the up and down that is so tightly controlled by  the cortex is lost and that's what's shown here. So  and in this cycle actually we're actually showing  the gamma has gone out. So something has completely  switched where the relationship of wake and sleep  gamma oscillations, the high-frequency oscillations,  is not only lost, it has reversed in certain cycles.  And then we gave, like I said PMP stands  for perampanel which is the AMPA receptor antagonist.  So this was our first pilot study and and we  gave a low dose which is 2 milligrams per kilogram  and we just gave two doses. So this is acute  dosing. Not long-term dosing to this mouse.  And then the same mouse, this P120 mouse, when he  got perampanel, it seemed like it started transitioning  again normally. So here you can see this gamma  and instead of going up in the same mouse it's  actually going down. So now gamma oscillations  have returned and that's why we call this as  perampanel-mediated rescue of the sleep/wake gamma  transition.

We were very excited to see this  because now this is hypothesis driven right. We hypothesized that since Rick's lab has  shown that AMPA receptor insertion is one of the  critical functions of SYNGAP1 at the synapse  that when it's not working and if there  are excessive AMPA receptors what could happen  to this transition and now if you give a drug at  a low dose that blocks those AMPA receptors then  the cortex now starts oscillating. Just like in  control. So then we went in. We went and quantified  this. So this is basically the wild-type like I  said the dotted line is form wake to non-REM which  would be that the gamma oscillations  are dropping and the solid line is from non-REM to  wake is when the gamma should be going up and  so you can see here on an from the negative  scale towards the positive scale and that's zero  line as you would expect all the way to non-REM  which is the dotted line or in the negative which  means the gamma, you know, gamma power is falling  these are slopes of the transition and then from  non-REM to wake, when the cortex is getting engaged  in gamma is going up, in the wild-type you can  see all of those slopes are in the positive  side and those two transitions non-REM to wake,  are significantly as expected different from each  other. What happens in Syngap this difference  is completely lost and so that significance goes  away. So whether it's one transition or the other,  the brain cortex does not seem to be recognizing  that it needs to switch to higher or lower gamma.  When we gave perampanel, acutely, the same mouse  we brought it back to quasi, very significantly  back to control. So what we can take away from this  data is the AMPA receptors seem to be playing  a huge role in these transitions and that a drug  that blocks those receptors can acutely switch it  back to normal. Okay. And this is a very low dose, two  milligrams is a low dose of perampanel.

Okay. What  are we doing for time? Oh my god I almost  took up the whole hour.

Mike: No, this is amazing please  keep going.

Dr Kadam: Okay. So then of course I showed you  the first Syngap child's EEG and we saw  seizures even happening between non-REM and REM.  So we wanted to look at those transitions. So we  looked at from the yellow to the blue and the blue  to the yellow. Now we will look at from the blue to  the red and from the red back to the blue which is  this and these are the heat spectrograms off the  EEG and I just want you to just focus on this is  non-REM to REM transition (this is in a wild-type)  and if you just eyeball it you can see that even  with a few second transition, the heat map  off that EEG is changing. Indicating now the  brain is transition from non-REM to REM. This is  from a Syngap mouse and that transition now  is not that self-evident. It's kind of blurred and  when you quantitate that by power and I'm going to  make this a little bit shorter by focusing on  here for the gamma because we've been talking  about gamma. The gray and the dark grey are the  Syngap mice and and this transition is very  weak compared to the wild type mouse and we give  perampanel (the blue) it kind of brings it back and  looks like the control. So these transition states  by perampanel are being rescued both from sleep  to wake and in during sleep from non-REM to REM.  And that's what we've done. We've given a low dose  perampanel. We've not done anything else and these are  adult mice.

Okay, so that is what I wanted to focus  on and we already talked about that. The other  thing we did was we did because I just explained  the importance of these PV interneurons and we  know their role in gamma oscillations and when  we had this finding but then beside we have the  brains from these mice and that's why we you know  we can take the sections, we can stain them  and we should look at PV interneurons because  there's lots of literature to show that PV interneuron or   the dysfunction, drives gamma oscillation dysfunction. So we  stained four PV interneurons which are in this fluorescent  green that's PV. GluA2 which is a AMPA  receptor subunit and then the blue which is just  a nuclear stains with every cell will show up all  the blue dots are all the neurons. The ones with a  blue dot and the green are the interneurons and  you can see they are far and few. So the very basic  ratio of interneurons to neurons in the cortex  is 1:10. So for every 10 excitatory neurons  there's one interneuron and that's all the inter-  neurons and then there's a subclass of those that  are PV positive. So this is just a binary image of  that on how we quantitate it and all these little puncta are the synapses that  these PV interneurons are making on all these  other blue cells. So if you just eyeball it you  can see there's a significant difference between  the wild-type and the Syngap mouse and so there  seems to be a huge reduction in how these  interneurons are branching, which we quantified here  and you can see the PV counts which are the puncta,  was significantly lower and the number of cells  themselves was slightly lower. For the GluA2 we  did not see any significant differences in the  sense of how many GluA2 AMPA receptors were  on these PV interneurons.

Okay. But when we looked  in the barrel cortex, so the somatosensory  cortex, we saw a completely different picture. Again  the color coordination is exactly the same:  the green on the PV interneurons, the GluA2  is the red, and because it's such a small signal  here is a magnification and the arrows are  showing you one specific interneuron (PV) and the  expansion is showing you that same PV interneuron  how much AMPA receptors it has on its soma which  is the body of the neuron and this is a wild-type  it has some, which we expect, but look at the SYNGAP1.  It has significantly higher GluA2 AMPA  receptors on the body of the interneuron, both  in these are layers of the cortex two  three and five six and this is iso cortical  meaning similar to human cortex and so we saw  this significant increase in GluA2. Now we gave  pure perampanel which is a drug which would block  these receptors and so the connection there  is PV interneurons we know have a huge role  to play in gamma oscillations and when we look  at the somatosensory cortex of these mice they  have this significant increase of AMPA receptor  expression on their soma. So this is what I just  said. GluA2 is generally expressed at very low  levels and in GABAergic interneurons and there's  previous work from other researchers and this  this group is in Germany, where they artificially  using genetic techniques did this right. It was  not a Syngap mouse. They took a mouse and  with genetic techniques did this to the mouse:  increased AMPA receptors was on to GABAergic  neurons. What happened when they did this the,  GAD-GluA2 mouse, it significantly disrupted  the gamma oscillation. So here is a mouse  that they just increased the GluA2 on  these GAD positive which is GABAergic neurons and  it completely disrupted the gamma oscillation. So  we are naturally finding this in our Syngap mouse  but when somebody else did this using genetic  techniques it resulted in the same disruption. So we  were very excited when we came across this paper.  

Ok so this is basically what I've been telling  you guys about, what Rick's lab has contributed,  which is a significant contribution and  this is Yoichi's work showing very elegantly  advanced and so this is a normal postsynaptic  spine here are the AMPA receptors in red these  little two things and there's a very very well  control cycle of how many receptors are at the  synapse, how they cycle, how they take it out of  circulation and how they re inserted. Ok. So now  SynGAP sits in this postsynaptic density and  when it is not functioning which is this greyed out  in this cycling gets disrupted and  then there's no control and more and more AMPA  receptors start getting inserted. Which is kind  of- when I showed you the thing  is this is what we predict is happening. So this  mouse has no SynGAP in the postsynaptic and for  some reason, specifically for these interneurons,  more and more AMPA receptors get inserted onto  the synapses in this cortex. So this had the  data that Rick's lab has shown matches what  we are seeing but specifically for interneurons.  So coming back to the circuit that you have seen  now if you put it into perspective that we know  these interneurons inhibit this neuron right  but we also know that these excitatory neurons  synapse on to interneurons. That is how they get  the signal. It's a feedback. When there's too much  excitation the blue circuit goes and tells  informs the interneurons okay you need to do  something and then they come and stop. So there's  input, the brain acts, the inter neurons stop the  signal. So that's how the circuit is controlled  very simplistically. Now what happens is as we have  seen work with the immunohistochemistry and with  Rick's data, what if now this excitation here  becomes fourfold? Which means this interneuron  will now keep getting signals to keep inhibiting  the pyramid neuron. This you can see is simple in  learning you would you would want some inhibition  but not constant inhibition. This would almost be  like a constant no-go signal and so that you  can kind of try and understand how that would be  for a cortex that's trying to learn something new. 

But how is this relevant to seizures? People have  done this experiment here. So what we just talked  about is exciting GABAergic neurons, exciting these  PV interneurons. So they use genetic techniques  and, you know, the people used when you introduce a  certain protein into specific cell types which is  here, these green cells which are the PV interneurons  and such that when you shine a blue light  on them they all start firing. Okay. So this is  a genetic technique and what they're showing here  is here's the blue light here are there are green  PV interneurons when you shine the blue light,  all of them start firing. They start firing. What  happens when you shut the blue light? So all of  these interneurons are giving a no-go signal  to all these interneurons 1, 2, 3, 4, 5. So that  is here. So when the blue light is on  they're firing altogether and we just saw how fast  they fire. When they are firing there's not even a chance  that any of these neurons can fire. So they're all  silent these five neurons. What happened when  they shut off the light, this group of these are  also in the group. Shut off the blue light,  this is called synchronous rebound spiking. Now  all the neurons... they fire together! They're not  supposed to! When they fire together what happened  on the EEG? The mouse threw a seizure. So this is  kind of reverse thinking where we think "oh, if the  no-go signal is weak that's how seizures come on."  What this group is showing if the no-go signal is  too strong the brain can also throw a seizure. Now  this concept has been around in the epilepsy field  for many decades but now with newer techniques  we are able to prove that this opposite way of  throwing seizures is the possibility.

Now why is  this important? I just showed you that there's  so much excitation onto the PV interneuron and  is that playing a role in these 3 Hertz discharges  that the cortexes of the children and the mouse  are throwing seizures for? So hopefully this is  the summary so basically what I've shown you  is we focused on cortical gamma oscillations.  These are these low amplitude  high frequency, and the wild-type transition high  gamma, no gamma. The Syngap Mouse, no transition in  fact sometimes it goes a little higher. You block  AMPA receptors, it reverts back right away. Back  to normal. So this is just the perampanel drawing.  It's an AMPA receptor antagonist. So when you give  the antagonist you can rescue that and these  transitions are associated with these 3 Hertz  discharges in the mouse. That's the same discharge  in the human. So that's the human EEG and that's  the graphic for that and then what I showed you  was the immunohistochemistry arm showing as  the way because it's the PV in wild-type and you  can see the red around here which is here's the  neck you know the black and white version of  it of showing how many more GluA2 AMPA  receptors there are on this interneuron from  the Syngap brain versus the wild type. So what we  this is denoting is that then this can become a  vicious cycle. This hyper excitation can lead to  this gamma oscillation disruption. This disruption  in transition can lead to seizures. Now do seizure  activity again lead to this further disruption in  excitation? That would be an interesting question  to parse out. So you've seen this before. Now you  know you're familiar with what this means for  the EEG of this child and so this is basically  you summarize the 3 hertz short wave spike wave  discharges. Seizures in transition states: wake  to sleep, sleep to wake, non-REM to REM. Early  morning seizures, which many parents report, the  clinicians have reported, and then that would be  a huge transition from gamma now suddenly has  to go out which then it fails to most likely.  Seizure clustering. Seizure clustering is  important because if they're not just single  events a bunch of seizures themselves can alter  circuits and disrupt learning and memory in any  child even in adult. So this basically summarizes  everything I have said: the non-REM component, why  we used perampanel. It was a low dose and how you  saw just with a single low dose how it completely  altered and that that is increased expression  of GluA2 receptors in these PV interneurons.  

So to summarize: SYNGAP1 patients I said 3  Hertz myoclonic seizures, the mouse has the same  phenotype. Then children evolved with multiple  seizure phenotypes we saw that too. Seizures in our  transition states, we saw that. Low sleep efficiency  we saw complete disruption of the sleep cycle in  the mouse. ADHD is reported in the syndrome  and then we saw the hyperactivity. What is novel, which usually  people don't look at when they don't look  at the background EEG in the clinical EEGs that  are done in the children, is this abnormal  gamma homeostasis meaning the transition between  states for gamma oscillation. Progressive worsening  of the epilepsy, which would be good to get from  the Natural History/registry that you guys are  working on and then this whole focus on GluA2.  Now why is this important? Perampanel blocks  all AMPA receptors. GluA2 is a specific  subtype of AMPA receptors and currently there's  no drug on the market that just blocks GluA2  specifically. So that would be interesting for us  to figure out if, you know, to try and either use  medicinal chemistry and screen for drugs that  would be even more surgically like a scalpel  rather than a hammer to block only certain types  of AMPA receptors and then maybe we could get away with  the side effects. Our future goals is where you  know we are collecting EEGs, overnight EEGs with  children who have been diagnosed with SYNGAP1  and then of course we're looking for different  ages because we want to see how this is evolving  in children as compared to what we've learned from  the mouse. We are now also, we've amended our IRB  study to include siblings. If there is a sibling  who has or if you can get an overnight EEG for a  sibling who does not have SYNGAP1 now that  would be the best control for that background  genetic background family for us to go to compare  sleep cycles with the sibling versus the child  with the mutation and of course we will keep  testing novel interventions in the mouse that we  have characterized but like I mentioned before we  are now move to a much younger age and we're doing  EEGs in pups. So here is the this is online and I  will I will send these links this is the proof  study at Hopkins and so basically it involves  parents signing a consent form that I will send  and if you have any questions like I'm happy to  answer them and then what it requires for the  parent to do is then sign a release form to  whichever clinic or hospital the overnight EEG  was done and then the hospital directly sends  us the actual recording not the report we are  happy to read the report but we want the actual  EEG recording the 24 hours or however long it  was if it's ambulatory EEG done at home we are  happy to take a copy of that too. So that is the  study we have been recruiting we have a few EEGs  right now and now we're focusing on analyzing it  now that we have this data from the mice and some  understanding of what this EEG looks like and with  that I'll close and acknowledge many Brennan who  was is the first author on the paper and did the  bulk of the work of the data I've shown you today  and then all the people who have been in the lab  who have contributed in multiple ways are funding  and of course Rick's lab that we wouldn't be doing Syngap  research if he had not invited us to  join the team and with that I'll thank you for  your attention.

Mike: Thank you Shilpa. If everyone wasn't  muted I think you hear a lot of clapping and I  apologize for being familiar right. I'm so humbled  by what I just see I feel like I should call you  Dr. Kadam more often but I've just been lucky  to talk to you in the past so please forgive the  familiarity. I know but you're awesome so I want to  just make three points and then I want to point  you to some of the questions in the chat.  Okay. The questions have come from Hans and Marta  who are both medical doctors and Syngap parents  which is as well as two questions from Aaron who  is the medical training as well and so they were  able to keep up with you. I think the rest of us  are still digesting but point one: you  said sibling EEGs. That's actually not  as hard as it sounds if you go to someone and  say "hey, can I get a sleep study" right because a  sleep study includes an overnight EEG so if my  chubby little two-year-old is having bad sleep  I can get someone to ask for a sleep study and  there's a sibling for you.

Dr Kadam: But right but I mean  it was are the parents of Syngap children who  said that their neurologists for some reason and  I'm really not clear on this, asked for or maybe  because if they were twins they asked for the  sibling the sibling without SYNGAP1 to have  an overnight EEG and as a scientist that could  be the because we would you know exactly what you  said most of our control EEG is because we will  have to age match because in children sleep cycles  evolve. A three-year-old sleep cycle is not the same  as the eight-year-old and so we are going to be  getting our control EEGs from the sleep center  at CHOP, Children's Hospital in Philadelphia and  those kids go for this exact same reason you said  that they have some of the child has some sleep  problems no genetic disorders and then they get  a diagnosis of no problems on the EEG. So we have  a collaboration with the investigator there and  once we have our cohort we will get age match  ages from that Center to kind of compare blindly  the Syngap EEG versus the control each year when  I became aware that parents did have sibling EEG  I actually took the trouble to amend IRB study  so that we can also get a copy of the sibling EEG.  Will it be very critical? I don't know. Once we have  a cohort and we are able to look at it it'll be  very interesting to look at the background EEG of  the sibling because that would be the best control  instead of a child unrelated child with maybe some  other disorders that might affect that  child EEG but that's the best control we can get  so whether it's easy or not is you guys will  give me feedback on because I think it happen  organically and I found out it's happening I don't  know how easy or difficult it'll be for you to go  and actively seek sibling EEGs for your  children.

Mike: I'll let you know but the second point  and I think the takeaway that anyone who  enjoys this presentation will walk away from this  is they're gonna do what I did after I read your  paper and walk into my neurologists office and  say "hey, could we try perampanel?" and my neurologist  says well, there's a lot of side effects and blah  blah blah and then I say yes, but what this paper  suggests (and please tell me if my interpretation  here is wrong) what this paper suggests is that  actually perampanel at the dose that it's  currently recommended for is an AED yes, we know  has side effects, but this paper suggested  at very low levels where maybe we could avoid the  side effects perampanel could have an impact  that's my first question to you my second one is  I just want to draw a thick line under something  you said earlier where you pointed out that the  same drug at the same quantity at a different  time of day yes to the seasonality or whatever the  right term is of our kids yours can have radically  different impacts and I've seen that just in one  of the sleep through that we give our kid guanfacine  for sleep and moving that from but we gave it to  him for behavior but it turns out to help sleep  we move that from morning to evening everything  changed. So I just want to emphasize that point  you made because I've lived it but the question  really is about the perampanel. Is it true when  we're talking to our neurologists about this but  we are talking about radically lower dosing that  is currently indicated at for perampanel as an  AED is that right?

Dr Kadam: A hundred percent. So the current  protocol for introducing perampanel to any child (or  any adult) is to start with the low doses and then  they'll ramp them up over a few weeks and then it  goes to they may not even start at  2. If they start at 4 the goal is to reach  up to 12 milligrams per kilogram. So that is a  substantially higher grade of perampanel dosing. But  the major point is: if you talk to any parents  I've talked to and very few Syngap children have  actually got perampanel (that I'm aware of) but other  syndromes where perampanel has been tried this is  the common complaint that there are multiple  side effects. But there's a distinction: there are  no side effects when the low dose is on. The side  effects come on when they ramp up the dose and  then usually people will just stop it. Right. And  that's understandable if you do it in an outpatient  and you tell your clinician down and now this  there's a child sleeping a lot or whatever the  problem is with the perampanel dosing.  So that's a consistent story. What our data is  saying because ours it is so specific let's say,  you know, these interneurons are showing this high  AMPA receptor, right. So what is the logic behind  that? Why is the low dose working or why we  think it's working? That only with the low dose.  The side effects come when the high dose block  most activity right? That's where the sleepiness  comes in and all the side effects come in.

Why is  the low dose working here? There are few cells as I  just showed you that have this very high level of  AMPA receptors. So if you gave a low dose that  same brain what would that the maximum effect of  the low dose would be on the cells that have the  highest AMPA receptors. Not on all the other cells  and so the understanding would be that you could  get away with the low dose which has no side  effects but have the most highest therapeutic  effect on the cells that have abnormally high  level of AMPA receptors and remember Syngap is  the only one that I'm aware of as a development  disorder where the specific pathology of high  AMPA receptor insertion is understood. Thanks  to Rick's lab. Most of the other epilepsy there's  in general high excitability which  they given perampanel to block the entire brain  from seizing so here perampanel would almost act  like a scalpel instead of a hammer it could be  for us it would be even better if we found a drug  that was only GluA2 specific which does not  exist (but might in the future) and the idea is  absolutely right then that this paradigm does  not exist for neurologists right now they only  understand the start to the low dose and ramp up.  

For a me as a scientist it's very clear that  the side effects are coming when the ramping  up happens not when the drug is just given in the  low dose. So that's one thing to consider. And two: if  the seizures are coming because of the disruption  of gamma oscillation then maybe (and we will test  this out) maybe just a low dose perampanel is  enough for the Syngap child. It may not be true  for any other syndrome it's very specific to  the SYNGAP1 mutation. The second important  point you said was when is this drug being given?  So right now Syngap is given you know usually  twice a day so you usually with an 8 hour difference,  sometime in the morning, sometime before bed. What  if we as we look at more EEGs and we figure out  that these transition points are when the child  is most vulnerable and... now if you look at all  the transition points I talked about is wake to  sleep, sleep to wake, REM to non REM. All these are  now tied to the sleep cycle. So if the only  dose if it's given only once a day is given right  before sleep and then your biggest side effect is  sleepiness then you counter both. It's a win-win.  And then you don't give it in the morning... and this  is all hypothetical right now okay, so I'm just  talking about the fact that what we are seeing  and how the drug would be best given. This has  to be parsed out in the clinic. So those are the  two times yes the time point when you're giving  the drug. What are is known side-effects? When are  you expecting the most seizures to happen  in the child? And then why would you give it in  the time when it's not needed? So those  are the points of things that I'm hoping as we  start talking about this data with neurologists  is changing paradigms and protocols in clinic is very  difficult. People do not... and I'm not even aware how  this ramping up dosing protocols came into  place like who tested it out how was it tested  out and you may know that Takeda now is actually  testing out the use of perampanel now in children  less than two years of age for safety. As a safety  trial. So it's moving in the right direction. These  two points are very critical. Why would you ramp  up if, based on the pathology, a low dose would  be actually the best way to go? And then secondly,  why would you give it multiple times  a day if you want to just focus on the most  vulnerable time when the brain is seizing? So those  are the two things.

Mike: Thank you. I know we're over  time but people aren't going anywhere because I think they're also enthralled. Have I forced muted  everybody? Should I unmute you Martha and  Hans and Aaron so you guys can ask your questions?  Here Hans I'll unmute you. I'll try to - it's not  working. Martha are you muted?

Marta: I'm here, yeah. I have  to use a different one because the connection  is bad. My question is: did you notice any  difference on the behavior of the mouse after with the Fycompa? Did you notice that they have the activity improve or something?

Dr Kadam: So that's what we're doing now. We're  actually... we are going to start looking at not  only how their behavior changes but how their  behavior is behavior testing for cognition  sure what we're going to see but now we know  what we expect to see now questions to be asked.

Mike: Thank You Hans, Aaron do you want to chime in  with your questions like one to two seconds  with like a 3.5 to 4 Hertz frequency oftentimes  the colleges are you observing those same kinds  of patterns in the mice and then also would you  classify those as seizures as well?

Dr Kadam: that's a very  good question and and that brings us to the point  of just animal modeling so that's why I mentioned  what the similarities were but obviously a mouse  is not not a child and we do when I said short  duration myoclonic seizures they are sharp but  they are not like 1 or 2 seconds long. We do see  one or two second long seizures that are not  the typical spike. I showed you the one where  the EMG was completely silent. So in that sense  it's not the exact phenotype which by the way  there is no such mouse model of epilepsy where  you have the exact human phenotype especially  not for genetic epilepsy. The closest we get  is with temporal lobe epilepsy where there's  very high seizure frequency and there's  only one type of generalized tonic-clonic  seizure but coming back to your more important  question is when neurologists, when there's a short  duration discharge and that's not classified as  a seizure that's just the nomenclature and it's  important for classification in the clinic and the  reporting but as a neuroscientist, any epileptiform  discharge we know, even a single spike, people have  shown disrupts the learning memory circuit in the  hippocampus. So whether it's a one second or two  second discharge versus a four second discharge  for me as a scientist that's still important to  identify and quantify and human EEGs really do  not go into that level of granular... it's almost ironic that  we use much higher resolution and granularity  when we quantify the mouse EEG but the similar  thing does not happen even with these overnight  EEGs when there are multiple seizure  events. Like, for example, the work, the  quantification and the trace I showed you for  the one child. The report did not say how many  events happened or where they happened. We had to  go and identify them and clock the time and  superimpose it on to the underlying power and the  sleep EEG. So basically to shorten that answer: yes in  the clinic a short duration epileptiform discharge  is not classified as a seizure band but for a  neuroscientist if a normal brain is not having  that kind of episode we know for a fact even  short duration discharges can disrupt circuits  and that's why it's important to us.

Mike:Amazing. Hans?  

Hans: Can you hear me okay? Yeah. Great, thanks so much  for the great talk. I just had a question because I  believe Gavin and maybe Jimmy had a paper showing  a diminishment in interictal spiking when SynGAP  expression was sort of turned on I believe with a  floxed Mouse. Is interictal spiking something  seen universally in our in our children or is it  something hit or miss?

Dr Kadam: That's a very good question.  So we have not started to quantify just spikes  in the EEGs. We are still collecting. Once we have a  good cohort and that's why we are urging parents  to send us, or participate in our IRB study.  For sure the report does not tell you how many  interictal spikes there were. Even in the one  that- so far the ones we've seen I have not noted  a lot of interictal spikes. Yes, we've seen the seizure  events but although they are the 3 Hertz  seizures that I was trying to correlate  between the mouse, they're not exactly the same  and the short duration high frequency 3 Hertz  spike wave discharges that are commonly  reported even in the reports are very unique but  they're not unique to send get and I don't know  even from the literature how many of when when  the epilepsy phenotype is predominantly 3 Hertz  spike wave discharges how many how what what role  do in trickle spikes play in that epilepsy and  that that literature is not that strong I would  assume it and it's never come my way and even  for the few EEGs we have looked at from the  clinic intraictal spikes of course they occur but  almost not at the frequency at which we see it in  the mice and this is I mean this is pre and this  is premature at this point because we have not  yet looked at multiple EEGs and so there it'll  be interesting to see whether certain children  who have the same phenotype but have a lot of  integral spikes that'll be very interesting so  both yes that is important point to look at the  ones we've looked at we don't have a very high  predominance of intrical spikes but that does  not mean that there aren't send your children  but there's lots of literature to show because  they are people are using Medtronic companies that  are implanting units onto the scalp of  patients is you are using a Gotham's to quantitate  interictal spikes to be able to predict when  that adult is most likely going to have a seizure.  So the idea being that then the patient will  be injected with the drug or the or the anti  drug into the focus of wherever the seizure is  starting and that's that's a new way of therapy  for in temporal lobe epilepsy or cortical lesions  so that's where interictal spike research is very  very strong. It's it's strong because it's being  used as a predictive marker for oncoming epilepsy  but in developmental disorders spikes do happen. In  Rett syndrome in their EEGs and they always  come in non-REM. What they mean for seizure  prediction is an interesting question that I  don't think people have yet started looking in  very young children. Does that answer your question?

Mike: He's muted. Yeah. Hans?

Hans: It sounds like it remains  to be seen based on a survey of child, Syngap  child EEGs, whether or not they have comparable  interictal spiking.

Dr Kadam: Correct. Correct. Okay great.  And in Rett syndrome we know there's massive  interictal spiking. I'm not seeing that- yes  once it's characterized in the clinic that'll be  very important. It will give us a lot of important  insights.

Hans: Great, thank you.

Mike: And so just conscious of time, next steps we're gonna  share out the recording of this pretty quickly  and then Dr. Kadam, if you can share the link to that  study you're doing we will make sure  to push that out as well.

Dr Kadam: Yeah I'll send you an  email right there this is done both I'll attach  both the forms and you can review them if you  have any questions please feel free I'll send my  email address too and and we are, we are trying to  collect these overnight EEGs.If you have  multiple, in the sense that your child  when they were three they had one and then they  had one again when they were six that would be  priceless for us because then we would have a  temporal progression EEG for the same child and  then if you guys have a you know one if you guys  talked you know and figure out whether sibling EEG  is something global please let me know and also  let me know how I can help.

Peter: Dr. Kadam, just to  follow up on that question, if we have an EEG from  after the child is two years old and one from just  before they're two years old overnight EEG would  that still beneficial? Yes it's only unfortunately because  of how the IRBs were initially written and  special forms need to be filled when the child is  less than two years old I will have to write the  amendment to get an EEG that's from a child  is less than two years old. So right now we are  approved from two years old to 16 years old but I'm  working on the amendment so at this point if you  have one that's over two years old you can send  that to us and then once the amendment is  done the consent form will already be  with us I'll just send you the updated amendment  and then you can send me the earlier EEGs.Sounds good. Which is, I mean I don't know but it's very  interesting for us as we're finding that there is  one subgroup of children that is having seizures  and severe serious very early on and the registry and  the clinicians who are working on this better they  characterize this the better it will be for the  field moving forward because definitely something  that's happening in the brain that's less  than two years old it's a whole different  ballgame then something is happening at five  years old and of course that the class of drugs  changes because most drugs are not approved to be  given for that younger child. So then that's why  management becomes different and then so then  the science needs to match what's happening at the  bedside and so then we have to start thinking that  way. The more we know the better we will do.

Mike: Marta any other questions  before I wrap it up?

Marta: My question is about the  age. With my daughter she's 17 she's one of  the oldest that then connected then you said it's 16 years old...?

Dr Kadam: So if extending it beyond 16 is not a  problem because out there there are no extra rules  and regulations with the two-year-old child we  less than two years old we have to fill these extra forms so now if I'm going to do  the amendment is that the only EEG she has is after  16?

Marta: No no no, she has, I actually had several of them  to send you. The other question is: she's probably  one of the few that doesn't have any seizures during  EEG. Or at least reported.

Dr Kadam: And you see...?  

Marta: She doesn't have any activity  reported and then but it still theta. EEG is abnormal  they're not...

Dr Kadam: 100% so I mean those are some  of the most valuable EEGs because we know that seizure  events also have effect on these background  oscillations. So if there is a recording of EEG  where it is not a single event that will also  provide additional insights and also like I  said, these seizure events most likely also are  dependent on the age. So there may be ages where  the child is more vulnerable to higher seizure  events and there may be "honeymoon periods" as have  been shown in other development disorders and we don't  know what that is for SYNGAP1 and what is  the range or whether there are specific stages  and whether all children fit in that stage or  there are mostly outliers. All of this information  hopefully now will start trickling in as more and  more people start getting involved.

Mike: Can  I...? So I'm gonna wrap it up unless anyone's  got a burning question. I want to thank everyone  especially the people from the Kadam lab for their  time. Do you mind going back quickly to slide 8?  I just think this is... I thought that was a super  cool slide but I didn't want to interrupt  you. That diagram is amazing because I think that's  what we live through right? We have young kids the  P60 who were only having myoclonic seizures and  the parents aren't sure and they go to doctors  and they go to neurologists and they take videos  and people say they're not sure and it takes time  and then a few years later, proxied by the P120's,  we have kids who are sure having seizures and  we start seeing different phenotypes. So point  1: I just think this slide really as a  Syngap parent and someone who talks to Syngap  parents every day really speaks to me and I  think it's awesome. That's point one. Point  two is I just want to highlight for the  parents generally, this is what we- this is why SRF  exists right? To connect parents and scientists. To  emphasize the great work that scientists are doing. There's a lot of names on that last slide but Dr.  Kadam has been someone who's taken our calls and  talk to us many times. Brennan who's the first  author on that paper is someone who was at our  round table last year and as a young researcher  and someone we're excited to see stay in Syngap  for a long time. So this has been a great  webinar and I'm so grateful and the third point  is just to build on that and say thank you. This was great. This paper's amazing.  Hopefully we'll have more kids on perampanel. I think  offline will ask you about why we wouldn't just  do a study. But thank you so much for the work  you do, for your lab, for this paper and for doing  this webinar. Thank you.

Dr Kadam: Thank you guys and thank  you for the opportunity. I mean the feedback from  parents has been one of the biggest motivation  for my group for Brennan for sure and and yeah  and for us as as researchers getting inputs of  what's happening in the life of the patient and  even in the family because quality of life  is an important component of us understanding. You  know "cure" is one thing but there are many ways in  which you can improve categories of the disease  and so although we can keep cure on the horizon  as the goal where we are headed to, we cannot miss all  these side routes where we can make substantial  improvements in the management and understanding  of how the disease is progressing and that  only comes from communication so thank you  guys for this. Thank you so much that was very good.  Thank you I appreciate it. Thank you. Bye bye guys.