SYNGAP1: The road from gene discovery to targeted therapy
Here are our introductory comments:
Marta: We are excited to continue SRF webinar series. The goal of the series is to get you closer to the science, making you aware of the
research that is being done and the opportunities to participate, and empowering your communications
Dr. Heather Mefford is a physician scientist whose lab is dedicated to gene discovery and pediatric disease, with a primary focus on the severe pediatric epilepsies. According to Dr. Mefford, in order to improve our understanding of the genetics of epilepsy and develop new therapies, we must go after the genes. She talks about how mutations and incorrect gene splicing can cause sequencing errors. When her lab sequences genes, they have to evaluate whether a genetic change is disease-causing or not, a complicated process because genetic changes occur in everyone and not all changes are harmful. She then speaks about the SYNGAP1 gene which was identified in 2009. Since its identification, research has shown that loss of function mutations in the SYNGAP1 gene, meaning that the SYNGAP1 gene is not performing its function, have led to intellectual disability and epilepsy in many SYNGAP1 patients. She finishes the webinar by talking about ASOs (antisense oligonucleotide), gene therapy, RNA approaches, and CRISPR as emerging therapies to combat limited or incorrect gene expression in patients.
Other Relevant Publications by Dr. Mefford
Today's talk is "SYNGAP1: The road from gene discovery to target therapy" by Dr. Heather Mefford. Dr. Mefford is an assistant professor of pediatric and at the University of Washington in the division of genetic medicine and attending physician at Seattle Children's Hospital in the genetic medicine clinic. Dr. Mefford's research laboratory is devoted to the discovery of novel genetic and genomic causes of pediatric disease. A major focus of the current work is to identify causes of pediatric epilepsy by employing whole exome sequencing, target gene panel sequencing, and custom array comparative genomic hybridization. She helps to create better diagnostic tools and treatments for patients who have health conditions with underlying genetic causes. The Mefford lab has discovered numerous new epilepsy genes. Dr Mefford has also been involved in the discovery and characterization of several new genomic disorders her clinical interests parallel her research interests and include seeing patients with genomic disorders and patients with severe epilepsy and neurocognitive defects of unknown etiology Dr Mefford relies on precision medicine to provide the most useful treatment options for patients and their families.
Dr Mefford: Great, thank you Marta, that was wonderful. It's great to be here today and I'm happy to give you kind of a a little bit of a whirlwind tour about genetic epilepsies in general, so how do we find genetic causes of disease which has been the focus of my group for about the past 10 years and then I'll focus on kind of what we know about SYNGAP1 and how that was identified what we know about the mutations and then give just a general overview at the end of emerging new therapies which hopefully some of them will apply to SYNGAP1 in the near future perhaps not all but I want to just give you kind of an introduction into the types of therapies that are under investigation by lots of different groups not necessarily mine but I know just enough to hopefully answer some of your questions about that and I'll just get started maybe there we go.
So as I said my lab is really dedicated to gene discovery and pediatric disease so we're not purely a SYNGAP1 laboratory we're more interested in identifying genetic causes of pediatric disease because we believe that it will be useful in identifying mutations in affected individuals mostly children and the disorders that we study and that we and others can work once we understand those genetic causes to understand the underlying biology the pathways the places that we could intervene to develop new therapies and improve the lives of affected individuals and what I like about the work that we do is that for the most part when we identify novel causes of disease, genetic causes of disease, this goes actually directly back to the clinical setting and gets incorporated into the test that probably for the parents on the call probably many of your children have a genetic test that included SYNGAP1 and every probably every month or so there are new genes added to these tests because of the discoveries that are being made at a rapid pace which I'll show you in a minute. As a geneticist I like to think about, you know, why do we care about getting a genetic diagnosis and I have to make this argument to insurance companies more often than I'd like but I think, you know, understanding of genetic diagnosis in a in any disease helps us on the clinical side give better counseling to families about prognosis. So what my life or my child's life going to look like? It helps us discuss recurrence risk which is, you know, will this happen again in my family? Either my immediate family or extended family? And in the case of epilepsy, among other disorders, it can affect the choice of medications, which we'll talk about towards the end, and provides research opportunities. Once you have a genetic diagnosis you can find those investigators working specifically on that gene disease or pathway and get involved and importantly it connects families with the same genetic diagnosis. That should be no secret here as SRF is a great example of this but there are many and I think over the past 10 years the emergence of family support groups and foundations and research funds through digital connections that's been one of the biggest advances that I've seen outside of just the basic science. And then of course the other reason that we do this is that we hope that understanding the genetic etiology will push us towards developing new and better precise therapies because it gives us specific targets to go after.
So this is an early view of what the genetics of epilepsy looked like 10 years ago or more and I'm just showing you kind of a very general discovery curve I guess. The first gene for any type of epilepsy was identified in 1995 and then there was this slow and steady uptick in genes identified with various types of epilepsy and the red circles here are more severe pediatric epileptic encephalopathies. And these were all, every one was a significant advance in understanding genetic causes of epilepsy. But we kind of plateaued and it was around between 2005 and 2009 that some new technologies were introduced into the field of genetics and genomics and we decided to apply these technologies to understand epilepsy and with the goal being again let's understand the genetic underlying genetic causes of... (I'll explain the "EE" in a minute) developmental and epileptic encephalopathy, with the goal to improve diagnosis in the clinics, genetic counseling, and eventually treatment. So I'm going to take a quick step back. For some of you this will be very basic but I want to make sure that those of you who aren't geneticists understand how we do this. So thinking about how do we improve our understanding of the genetics of epilepsy? Well we go after the genes, right, and because we're going to be talking about genomes, genes and DNA and genes I'll remind everybody that when I talk about the human genome that's really all of the DNA. So every single cell in our bodies carries 23 pairs of chromosomes, those little XS here which are packages of DNA and you can think about this as kind of the full instruction set or information set for building and maintaining our bodies. Each chromosome is a packet of DNA really and you can think of that as one volume of the encyclopedia. And on each of those chromosomes resides a number of genes which are the basic construction units and they're kind of scattered across the chromosomes.
So not all DNA encodes a gene and if you think about it if our encyclopedia analogy are these readable sentences here in red among other DNA sequence that is not as easy to interpret and we'll get back to that a little bit. So think of them as short sentences or short paragraphs that provide the actual instructions for making proteins and again building all the all the parts of our body. So humans have 20,000 approximately individual genes. In each of those genes is a code to make a protein and we can read that code. And the ability to read that code is important because that's how we know when the code is messed up and provides incomplete or incorrect instructions. And so you can think of that as a spelling error or "typo" which can happen when we copy DNA. Our bodies are very good at copying DNA but there's so much of it. Three billion individual letters and so there are always mistakes that are made when we copy DNA. Most of them don't have a noticeable impact on our health or development though. So you can think of it like changing one letter in a sentence so if you change a single letter it can significantly change the meaning of that sentence. Or it can have a very subtle impact or no impact on the meaning of that sentence. So this is an example here of sitting in the context of the gene for sickle cell disease and again one single letter among thousands causes a change in the sickle cell gene that results in sickle cell anemia. So these types of genetic changes, which you can think of as spelling errors are the missense, what we refer to as missense changes, but importantly you can also have genetic changes that just stop. Basically stop reading, right. These are, we refer to them as premature truncations, nonsense, or stop mutations and these are the types of mutations that are actually very important in Syngap. The other types of mutations you might hear about all have the same, usually have the same effect of prematurely truncating the protein include frame shift variants and splice variants.
Splice variants are important because genes are not actually just a single readable sentence in our genome and they're broken up. And so you can think of sort of each of these exons as one of the words in the sentence that need to be sliced together and that splicing machinery takes out the intervening sequence and puts the puts the gene together in a way that encodes the protein and this is the code that we actually read. So if you mess up splicing, for example you might not include this middle exon and then again it takes out a chunk of sentence and it causes an error.
So how do we find these sequencing errors? Well we read the DNA and for a long time we had a very robust (but slow) sequencing method. This is a traditional method. The actual output is shown here on screen by these red, green, blue, and black traces here that read the DNA. And it was in about two thousand, I don't know 2008, 2009, where we finally developed a new (not me but some brilliant people) developed a new sequencing technology so "Next Generation Sequencing" or "high throughput sequencing" where we could actually do this much, much, much more quickly, much more cheaply and sequence lots of DNA in a short period of time. And just to give you a sense for what an impact this has had on this field of gene discovery in human genetics: we sequenced the first human genome so this is the Human Genome Project using this first technology in about 10 years for 3 billion dollars and today using Next Generation Sequencing I could sequence the genomes of everyone on this call in a matter of a few days for a thousand dollars each or less. It costs a little more to actually and takes a little longer to interpret all of that information but this has had an incredible impact on gene discovery. And so our group has used these technologies to study specifically the severe early onset epilepsies also known as Developmental and Epileptic Encephalopathies and other related neurodevelopmental disorders. And as you all probably know these are fairly severe conditions where often kids will have seizures in the first days, weeks or years of life and they'll have associated developmental delays and sometimes poor outcomes and this can be developmental outcomes or health outcomes but it has a significant impact on quality of life. Individually there are many different types of Developmental and Epileptic Encephalopathies both from kind of how they present from a clinical standpoint but also from the genetic etiology and each individual DEE is rare but collectively they start to add up and they're actually more common than you would think.
The seizures are often intractable in these disorders and despite all of the medications at our disposal which is over 40 anti-epileptic drugs today, many of them are not effective even in combination for these kids, speaking to the need for more precise therapies. So we have over the years collected over a thousand kids with a DEE presentation, so severe early onset epilepsies and developmental delays and by "we" I mean not just my group but we collaborate with some wonderful researchers in Melbourne Australia, Ingrid Scheffer who many of you have probably heard of, with our collaborators locally and regionally as well as people that we meet across across the internet at meetings with people across the world who refer patients to us. And then we use the sequencing technology to look for the genetic cause in those individuals and we can do this in very similar ways to the clinical tests that are available today. So we often start by a gene panel. So sequencing 10 genes or 50 genes or 100 genes at a time or we can sequence all 20,000 genes which is exome sequencing and we've used both approaches for different reasons at different times and I can answer questions about that afterwards if you like.
So I think it's important to understand also how do we... we find genetic changes in everyone so every individual has differences in their DNA compared to the person sitting next to them and how do we decide if that genetic change is disease-causing? Especially when we're identifying new disease genes. And these are some of the criteria that we use and I can get into the nitty-gritty later if anyone's interested, but basically the genetic changes that we see in things like Developmental and Epileptic Encephalopathies we don't usually see in healthy people and we have public databases available now that provide information about whether specific genetic changes are seen in the population. As I said, we can read the code for every one of these genes and so we want to know that the gene is predicted to impact the protein, so have a damaging impact on proteins and there are different ways that we can discern that. When we're looking for new genetic causes of disease we like to see that multiple individuals with the same disease have changes in the same gene and then when we're talking about DEE, as many of you know, these changes are often new in the affected child compared to their parents and that's just shown here in this what we call parent-child trio where there's a genetic change here in the child that's not present in either parent. And importantly every single one of us again has these types of changes that are new compared to those changes that we inherit from our parents so we have to take those new changes in context of all the other things that I've shown here on this slide again to determine that it's disease causing.
So using this approach our group and many others who've taken similar approaches has greatly has greatly impacted gene discovery over the past 10 years and I'm showing you here just a subset of the genes that have been identified to cause Developmental and Epileptic Encephalopathies many of which are quite similar phenotypically to SYNGAP1-related disease and there are well over 70 probably 100 new genes identified (they're not all listed here) and Mike pointed out to me that SYNGAP1 is not on this list but I added it and importantly I think it's remarkable to realize that SYNGAP1 variants were one of the first kind of "new genes" identified and it was first reported related to human disease in 2009. And the first reports were not really focused on epilepsy. They were focused on the intellectual disability aspect and a lot of the early research in gene discovery in pediatric disease was intellectual disability. So this is the paper in 2009 that first reported mutations in SYNGAP1 and they sequenced 94 individuals with intellectual disability and found three who had a de novo truncating variant. They also sequenced some kids with autism, some individuals with schizophrenia, and 190 unaffected individuals for SYNGAP1 and found none of them to have variants in the SYNGAP1 gene.
So this was the first report and over the next about three years several additional studies trickled out reporting additional individuals with variants in SYNGAP1 and there were total, when we were starting to think about identifying causes of epilepsy, there were a total of about 18 patients who had been reported including the three in this first paper and we noticed that at least 12 of those were reported to have epilepsy in addition to their intellectual disabilities. And so we took the approach of saying well, we want to know what if we sequence a bunch of individuals who come into the clinic with a primary complaint of epilepsy who may also have developmental disorders or developmental delays. And so that's what we did and reported in 2013, was sequencing 500 individuals with an epileptic encephalopathy and here we took a gene panel approach. So we sequenced 65 genes. About half of those were known genes so already associated with epilepsy or proposed to be intellectual disability genes or autism genes and then the other half were genes that were not yet associated with disease and so we were testing the hypothesis that they would be a cause of epileptic encephalopathy. And so CHD2 at the time was not associated with any human disorder and we identified mutations in this gene and showed that it was a cause of DE but we also found mutations in SYNGAP1 as one of the most frequently mutated genes in our cohort as well and this was work that was led by my postdoc at the time Gemma Carvill and with important obviously contribution of patients from our collaborators in Australia. So this was what we reported at the time. There were again 500 patients that we sequenced and the bars in blue were genes that were already known to be DEE genes and so we found lots of variants in SCN1A, CDKL5, STXBP1 and then CHD2 and SYNGAP1 were the next most commonly changed genes and we found five individuals with de novo truncating variants in SYNGAP1 and concluded that this was a cause of not just intellectual disability but also could be even more severe Developmental and Epileptic Encephalopathy. So this is a diagram of the SYNGAP1 gene. This is also from this paper and the variants in red here are the ones that we identified and all of them are truncating variants. The variants in black were previously reported in the literature and those in bold all have epilepsy in addition to their developmental delays and intellectual disability and those that are not bolded did not have seizures.
So more recently this is work led by Ingrid Scheffer that I'm sure you're all familiar with but they looked at and described 57 individuals most of which had not been previously reported with variants in SYNGAP1 and they described both the genetic landscape as well as really nice detailed phenotype information. And I'm not going to focus on the phenotype today, I'm going to focus on the genetics, but importantly they found that 34 of the variants again were those truncating, so nonsense/stop/frame shift variants, eight of them were splice site variants which usually lead to premature truncation, four were deletions that included SYNGAP1. So one copy of SYNGAP1 was deleted and then only 11 of them were missense and so I think this starts to tell us something about the mechanism- oh sorry, here's a figure from their paper and it's very busy but what you need to know is that this top figure, again this is the gene. These are all the exons in blue and the truncating variants are on the top and then the missense variants are on the bottom and these are patients not in their study but but in the literature, and again truncating variants on the top and missense variants on the bottom and what you can see is that first most of the variants are truncating and second the missense variants tend to fall into these colored domains. Not all of them but most of them and these are regions of the protein that we know to be important for its function. So the thought is that changing one amino acid in those regions may disrupt the function of the protein.
And I think this is important for thinking about what is the mechanism or why the variants in SYNGAP1 cause disease. We have two copies. These variants occur only in one copy of the gene and the fact that most of them lead to premature truncation, so one of the copies is just not made and not functioning, tells us that it's most likely loss-of-function of the gene or haploinsufficiency. "Haplo" being half. So one copy is gone and the other is normal and functioning but the amount of SYNGAP1 that's present is not enough for normal development. So loss of function likely the mechanism. There are lots of nuances to this that we can get into later but we generally think of genetic disorders as due to loss of function of the gene or gain or increased function of the gene which can sometimes be actually a new function of the gene and that does not seem to be the case for SYNGAP1. So as I told you we didn't identify just SYNGAP1 but over the years we've we've found the genetic cause in many of our patients and what's important to remember about neurodevelopmental disorders in general is that there are *many* different genes where a mutation can cause a similar phenotype and here just each bar represents a gene and the height of the bar is the number of patients in our cohort whose genetic cause is due to that gene and these orange bars are the genes where we played a significant role in first reporting variants in those genes associated with disease and then also in reporting kind of genotype/phenotype correlations for some of the newer genes. So if we think go back to why do we do these genetic studies and find genetic causes the reason is it helps us understand the biology so what what different things can go wrong and lead to abnormal development and then now we have specific targets.
So how do we use this information to develop new therapies that might be more effective and more precise for specific genetic disorders? So how do we do that? Right? The goal is when you get your genetic diagnosis there's information about specific treatments. So this is an area where there is not a specific SYNGAP1 therapy yet but I'm going to tell you about some advances recently in developing new types of therapies but I'm first going to tell you what I guess what what do we know about epilepsy and how precise can we get today so part of the reason I want to talk about the mechanism is it loss or gain of function or what's the pathway is: it matters. For precision therapy that's precisely why we do this and there are some examples of treatment recommendations for specific genetic disorders involving epilepsy and development. So one is GLUT1 deficiency, so SLC2A1 is the gene and this is when you have a mutation in this gene you have trouble transporting glucose or sugars into the brain for energy and so the ketogenic diet helps generate alternative fuels to get into the brain for proper functioning. So if you get a diagnosis of SLC2A1 the recommendation will likely be that you go on the ketogenic diet and it is effective.
Another example is an epilepsy due to mutations in this gene here ALDH7A1 and the important thing about this is that it blocks vitamin B6 production and utilization and so the treatment is to go around that blockage and treat the patients with vitamin B6 which stops their seizures.
When we think about gain of function and loss of function that also matters. So SCN1A which causes Dravet Syndrome, mutations in SCN1A cause loss of function or haploinsufficiency and SCN1A encodes a sodium channel. So if you're already missing, you don't have enough of the sodium channel, you really don't want to further block sodium channels and some of the drugs that we use for epilepsy are sodium channel blockers. So knowing that you have an SCN1A mutation your neurologist will most likely avoid sodium channel blockers because we know that these can make the disease worse in some individuals. But on the other hand there are two other sodium channel genes where the mutations that cause epilepsy are gain-of-function mutations. So this is a really important distinction because in these individuals the drugs here on the right, which are sodium channel blockers actually can improve disease. Now neither of these are magic bullets for these disorders but they're important because you need to know whether you're going to make the disease better or worse by using this class of drugs.
So that's great. This is taking information about genetics and using the medications that we have available on the shelf today to our best ability but what are the drugs coming down the pipeline? So I think there's some really really exciting emerging approaches and emerging therapies that I'll tell you about and some of these may very well be effective for SYNGAP1-related disease if we can find the right one. So there are some buzzwords that you may have heard in the literature or in research talks that you've been to. The first is "antisense oligonucleotide" or "ASO" therapy. So antisense oligonucleotides are really small. There's just 20 bases or so. Single-stranded RNA. So not DNA but RNA molecules and you can use ASOs, they're designed to go in and bind to the RNA. The RNA is the intermediate between DNA and protein and you can use them to bind to the RNA and depending on how you design it you can actually increase expression of that gene, and amount of the protein, or decrease expression of the gene and decrease the amount of protein. You can also affect that splicing mechanism of the gene. So you can go in and block splicing that shouldn't be happening that's happening because of mutation or you can interfere with splicing in a way that might correct splicing changes.
Gene therapy is another technology that has been around since I started grad school. There was a lot of hope for gene therapy. It's been a lot harder and longer road than I think we ever anticipated to have effective gene therapies across human disorders but there have been advances and there are promising therapies for some disorders and some that are on the market already. And then there are other approaches using RNA approaches to increase or decrease gene expression. The little diagram at the right is something called a SINEUP which you can design. Again it's an RNA molecule that you go in, you design it to bind to your gene of interest, and it increases the amount of protein that's produced from that particular gene. CRISPR I don't even have on this list but obviously that's in the mix as well. Thinking about how to use CRISPR to correct rather than just replace the specific mutations in the gene. And so importantly like I said there have been advances. There are some of these that are already available clinically. Spinal Muscular Atrophy is a very severe fatal disorder that now has great options for therapy. So there's an ASO therapy that is in use right now. Nusinersen is the name you might hear and there's a gene therapy for SMA: Zolgesma which is available for kids I think under the under the age of one of them is under the age of two one of them is even earlier but these have been actually very very effective and life-changing for many kids with SMA and there are probably other examples out there but these are the the two that come to mind as particularly effective. So what about epilepsy and neurodevelopmental disorders?
Well there's great I think promise for these types of therapies as well and the most recent and exciting example is for SCN1A-related disease particularly Dravet syndrome. In this paper we've kind of heard about these studies over the past couple of years but the paper was finally published just in the past month or two and this is an ASO for SCN1A and I'll show you exactly the mechanism in a minute but basically what you need to know is that the ASO results in an increase in SCN1A protein and so in the study of mice with Dravet syndrome when you give the ASO at day of life 2 in the mice what you see is basically that it prevents seizures and it prevents the premature death that inevitably affects SCN1A mice. So the purple line here is normal mice who 100% survive out to 90 days because they don't have an SCN1A mutation. The red line are mice with SCN1A mutations who didn't get treated and what you can see is that at about day 20 or so they start to have seizures and usually when once they have a seizure they have sudden death. And you can see that by day 30 about 80 percent of the mice have died. But when you treat mice with this ASO therapy, that's the blue line here, the vast, vast, vast majority do not develop seizures and survive out 90 days just like wild type mice and so this is very exciting. Humans are not mice. So we need to always need to keep that in mind and for most kids with Dravet syndrome we don't diagnose them on day of life two before they've had a seizure. We usually diagnose them because they've had a seizure and so their second study within the same paper treating mice closer to the time they develop seizures and it's still an effective therapy. It doesn't look as pretty as this curve but it's still very, very effective and so thinking about when to treat and how to treat is really important and what effect, what differences in effect, a treatment will have based on the timing is important as well.
For those who want to know the details, the idea here is that SCN1A actually has an extra exon. This pink exon here is just a cartoon version but it's what we call a "poison exon" because when you splice in this exon to the SCN1A gene, it actually introduces a stop. So it's similar to the stop mutations or premature truncation mutations and so the full length protein is not made and this is normal but it's probably normal in cell types where you don't need SCN1A or developmental time points where you don't need SCN1A. So the idea for this ASO is that if we go in and block incorporation of this "poison exon" then the cell is more likely to make more of the normal full-length SCN1A and that's exactly what happens. So the ASO goes in, it blocks incorporation of the "poison exon" and you have more full-length SCN1A.
This ASO therapy is now in clinical trials for patients with Dravet syndrome through a company called Stoke Therapeutics. So it'll be really exciting to see how effective this is in the human population. And they have treated their first patient. I haven't heard any outcomes. I briefly touched on SCN8A-related epilepsy earlier. This is where mutations cause a gain-of-function of the gene. So you make too much of a copy that's not working properly or actually you make a copy that's not working properly and increases the function of a channel in a way that's deleterious and there's a recent publication from a group in Michigan led by Miriam Meisler where they used an ASO to decrease production of the SCN8A protein. This is just the opposite of for SCN1A where they use it to increase production of protein. But here you want to "knock down" SCN8A because you have that copy that's hyperactive and so they actually treated mice again who had SCN8A mutation that you can find it's the same mutation found in some humans and causes disease and they found that treating with that ASO increased survival of the mice. It delayed seizure onset and it improved their motor activity. It was not, they had to get, so for the SCN1A ASO they actually just gave a single dose. For the SCN8A ASO they found that giving repeated doses over time was actually more effective than just a single dose. So again learning about how how to treat and how to use these novel therapies in the model organisms is going to be important. Interestingly they also treated mice with mutations in SCN1A not SCN8A but they use the SCN8A ASO and it improved survival in those mice as well. This is a little more complicated story but the reason I bring it up is that sometimes when we develop therapies for one specific genetic cause of a disease it actually may have broader implications for other disorders that are related or that have mutations in a different gene but that's in the same pathway. And so I think it's important to remember that maybe we don't have to develop a therapy for every single gene but that we can work together with within pathways or within disorders and hopefully to have targeted therapies that work more broadly.
So I'm going to end with thinking about you know the promise of these novel therapies is huge. It's really exciting right now but I think as always we need to be cautious and we need to think about a lot of different possible outcomes and I think for me when I, you know, talk with patients in the clinic about, you know, why can't we just do this? There are lots of things to think about. So again what is the right level of gene expression for a particular gene? So we have these techniques now where we can add a gene back or we can turn up the volume for that gene as a level expression we can turn down the volume but we've got to get it right. So genes like SCN8A, while there are gain-of-function mutations that cause epilepsy, loss-of-function of that gene can also cause movement disorders and intellectual disability. So, you know, we need to be careful about what if we turn it down too much do we actually cause a different disorder? What cell types do we need to get the therapies to and how do we get them there? These are disorders of the brain. The brain is a tough place to get to and, you know, how much do we need to give and how often to get into the right amount of cells to actually have an effect? How specific is the treatment? So again using an SCN8A ASO treated SCN1A mice. That's great but what if we have a treatment that has an unexpected effect on another gene? That again might relate in a different unexpected outcome that may or may not be beneficial. And then importantly what if there are side effects? Can we stop the treatment? Gene therapy is hard to take away. The ASOs, that might be a different story. In some of these other RNA-based treatments medications that we use today you can give the medication or taper it off if there are side effects and take it away. So we need to be careful in thinking about how do we do that for some of these novel therapies so that we're always having a beneficial and not a deleterious effect on the patients that we treat.
So I think I'll just end by saying I think there's been a huge, you know, increase in our understanding of the genetics of epilepsy. I showed you that there were, you know, 100 different genes and this has been extremely useful in the clinical setting and for connecting families and scientists to better understand these disorders and every single one of these wants a precision therapy and I think there are a lot of people out there working on it and there's a lot of promise. It's important to remember that we can use, you know, studies in in cell culture and cellular models, the model organism studies in mice, zebrafish and other organisms are going to be really important and they take time. And then many of you are probably engaged in research which is fantastic because using patient cells also helps us as well where we can use patient cells to study different types of cells using a stem cell approach that can allow us to make brain-like cells to better understand the disease.
I'm going to leave you with this again to remind you that SYNGAP1 is one of the early players. I think you know that means that actually people have been studying it for a long time before it was identified as a human disease-causing gene and this, you know, 10-12 years here has given us the chance to identify lots of affected individuals especially now that these tests are widely available in the clinic and to understand what is the phenotypic spectrum, what's the mutation spectrum and how can we use that to understand the disorder and develop these therapies. I think that is my last slide. I will stop there and I'd be happy to take questions.
Mike: Thank you very, very, very much. That was amazing. I think we're going to make this recording required viewing for every newly diagnosed family. Do you want to take down your screen just so we can put faces on? So, three things, Dr Mefford and then I would encourage everybody who's watching to just throw throw in questions in the Q&A. The first thing I want to say is thank you both for this remarkable and thorough talk but also for all your great work. You know your name was on those papers that helped identify this gene that our kids have so amazing and we're just so grateful and also for your work with us on the SAB. I have two questions one that I feel like I have to ask because it's probably top of mind for many parents and one that I want to ask because I'm dying to ask you. So the one that's top of mind is VUSs right. We have parents who got a VUS on their Invitae panel and then x years later it was resequenced or the phenotype presented in such a way and, like, how should parents think about VUSs? What does that mean and then sub question there what is the difference between an intronic VUS which which a few people have and then the phenotype gets so profound it's determined to be pathogenic and what would a VUS be that's not intronic? So that's like I think the question I have to ask for the Syngap family audience and I'd love to hear from you about that and then I'll... then the other... I'll wait for the other question because that was enough.
Dr Mefford: Yeah so I will tell you that the VUSs are as frustrating for us as clinicians and scientists as they are for families and it's sometimes the most difficult kind of, you know, visits to have when you have to give family a result that is uncertain. I think it's, you know, sometimes geneticists can take a report that says this is you know a VUS and we can say "no it's not. I know this is disease-causing" or "no it's not. I know this is not causing disease in this family" using lots of different tools and experience but often that's not the case and so when that happens, you know, I think it's important to engage with, if you can, with someone who knows a lot about the gene. Still might not give you the answer but to also over time remember to revisit these things because you can, you know, we do ask for reinterpretation of variants over time because interpretation can change. It can change because we've identified additional similar patients with the same genetic change and start to say "okay, this is actually, we think this is actually causative" or you find unaffected people with the same change and then you can say "we don't think this is the cause in your family". If you've had targeted testing it doesn't really happen very often anymore usually these are pretty big gene panels but even with a big gene panel, you know, you can expand that test to a whole exome and say "well, did we miss something and this is not actually the cause there's another cause?" and then sometimes there are follow-up tests you can do. So intronic VUSs, you know, the thought is that some of those might affect splicing. So we understand splicing to a certain degree but we don't understand it fully and so sometimes we can say "gosh, I don't know if this affects splicing or not". We're getting better at developing tools to predict whether it affects splicing. I still think it's hard to base your entire clinical decision making on a, you know, AI tool or, you know, a bioinformatics tool but it can be helpful and or you can go in and and do more experiments but again you have to connect yourself to a research lab that was willing, or sometimes you can do it clinically, to do these experiments to say does it actually have the effects that we think it has that would be a disease causing effect? And sometimes that takes years to be perfectly honest. So just know that it's frustrating. I think the hard thing too is that I will tell you as a geneticist the hardest thing that I ever have to do is take away a diagnosis and so there are cases where people might get a VUS and connect themselves and just engage, you know, in the family community and research community and with the belief that this is the cause and years later we might figure out that it's not, right. So I think just knowing that that's a possibility you know if you have a VUS but there's nothing else to go on, you know, connecting yourself to a community is important regardless because affected kids are affected kids and many of them have the same types of, you know, issues and challenges so connecting with families is important but knowing that that may go away someday I think can be really hard but it's important to have that kind of in the back of your mind. That makes sense?
Mike: Thank you. Yeah it does. It does. My message to anyone with a VUS is interrogate it more. Find an expert. Do an RNA test if it's intronic whatever...
Dr Mefford: I think the other thing to know is that for some genes, you know, people are starting to develop assays where we test every single change. So if you know... if pathogenic variants cause the gene to act in a certain way or not act in a certain way and you can test that, you know, then we can say "well, let's take your VUS and let's test it and see if it acts the way we think it should". You can also say "you know what, we're just going to do this for every single possible variant in the gene and then we're going to have a record of this variant as normal, this variant acts like a disease variant". Those are hard to do and you have to have the right test and blah blah blah but those those kinds of tests are starting to happen for lots of different genes so that we can proactively have the information rather than saying "oh, don't know. Give us the variant and we'll take a few years to figure it out" right. Yeah so that's probably down the road. I don't know that it exists for SYNGAP1 but someday it might.
Mike: So the second question and then I'll let you go to the Q&A where they're piling up, is about Stoke and SCN1A and SYNGAP1 in the flight path. So what I mean by that is, like, you know when when we have on our board Hans Schlecht who teaches me something new every day and one day he was like "SCN1A is the one. That's how you remember. That's Dravet" and because it's the biggest one and it was the first one and so I'm always looking to that to see what's next for us and what I try to say to parents is it's important we watch these other major genes because we're a half step behind them and I think Syngap really is a half step behind them because we have a large estimated prevalence, we're a classic haploinsufficiency and these are all very Dravet answers. Even in the case of Stoke, if you read their patent, that mechanism is for two genes SCN1A and SYNGAP1 but where we get off the path is, you know, thank god our seizures aren't as catastrophic as Dravet but it also makes it harder to figure out what you're going to test. Like, those mice died or they didn't. Right. Our mice don't die they just behave oddly and fail weird tests. How do patients think about that ASO versus other ASOs and what is... here's the question: based on everything that's going on with ASOs and Dravet and the speed of change how how hopeful should our parents be? I know it's going to be an impossible question like how many years do we get a therapy for our kids but should we be thinking months, years or decades? I mean orders of magnitude here? Like, what's what's a reasonable hope if we keep pushing?
Dr Mefford: I'd like to say months. I think realistically it's years. I think it's better than decades so I think, you know, given the time that it takes to do these experiments in mice, to make sure that it's safe, I mean I think it's a little off track but thinking about how we're watching the COVID vaccine development right. That is science in real time and now the whole world is watching and it's going to take time, right. The record to make a vaccine is absolute record is four years and now we're pushing it to six months to a year and you're seeing all the steps in the process and how that works and so I think just I only say that because I think it kind of gives us a sense of how science works and where what the challenges are and how fast we can actually move. You know it'd be great to have a warp speed program for ASO development right and I think we're getting there and I think having industry involved and engaged in rare disease research and treatment is huge and I think we can be hopeful and I do think you have to be cautiously optimistic because again all of the issues of when do you need to treat how soon or how you know what part of the disease process will it actually make an appreciable difference is important. It's important to be realistic about that and I think your question about what is it that you're treating is very important. So seizures are an easy end point to look at. Development's harder. Especially when you have a range of development. Behavior's harder but behavior is huge right and as you all know it's important and so I think figuring out what those endpoints are going to be and what order of magnitude of change do you want to see in treated kids versus those who don't get the treatment are all things that we have to be thinking about.
Mike: Yeah and I think having remarkable people like yourself lending your good name to our gene can't hurt when companies are making choices so thank you again. I'm gonna go to the questions. There's seven in there right now and I'm gonna start at the bottom with the names that I know. So Dr Schlecht is asking could you talk about the Angelman ASO if you're familiar with that and if you're not... I'm not sure I'm familiar enough to talk about it intelligently. I know it's out there. I know it's happening. It's a slightly different mechanism but it is an ASO therapy that is in development and I think again it's important to watch that space. And then J.R. is asking a question right above Hans is when will the data analysis strategies be ready to look at whole genome data as a way to find modifiers of disease? Who would do this for SYNGAP1? And then I'll give you her other... okay one at a time, go ahead.
Dr Mefford: No, no, I think the modifier question is really interesting and it's a hard question. It's definitely something we've thought about not just with SYNGAP1 but with some other disorders where there's really variable outcomes and I think the key is doing exactly what you're doing: identifying and bringing families together who are willing to participate in research because in my mind it's going to take numbers, right, because (unless you get really lucky) so think about modifiers right what could be modifying this? Lots of different genetic variants but to find it you have to have a concentration of variants in a set of patients with similar phenotypes right and to do that takes collecting lots of data on phenotypes and takes probably lots of patients to study on the either the exome or genome basis. So, you know, we can talk about doing that eventually and we definitely thought about it and again I think similar to the questions about treatment outcomes you have to think about what are the variables that you're looking for that you're modifying, right. The development of the seizures that you either... yeah.
Mike: Yeah. Million dollar questions. All right. So JR asked another question: since we know SYNGAP1 mutations are fairly prevalent in ID and epilepsy how do we go about finding older patients? This is a super cool question because by the way we just found a 65 year old by accident. Yes I heard. A doctor was real and we're making a movie about... but that's another thing. Most of us have young patients or parents of teenagers that push for diagnosis. What is the ethical way to go about this? I think "this" meaning finding older patients.
Dr Mefford: So I think there's a lot of movement in both the clinical community and research community to do genetic testing in older patients. So, you know, older patients didn't have the advantage of having this technology available or even knowing that Syngap caused you know similar phenotype to theirs so there are some good, couple good recent studies looking at older populations. I think the way to do it is to engage adult epileptologists and physicians in general who would see patients like this, right, to get them to do genetic testing which means getting insurance to coverage testing in that population and to engage researchers in finding populations right who... they're out there they just haven't had the advantage of testing. So thinking about, you know, you could go to group homes where people live. The challenge with this is getting consent for research and we've talked about that with one of our adult neurologists here who cares for a lot of these patients but often it's figuring out who gives consent, you know, for a blood draw for genetic testing? Who gets the results and what do you do with that information? So I think thinking carefully about how we get to those populations both on a research and clinical side is important. I think it is really important for families to be able to look ahead and say what are the issues that are going to come up in my kids when they're 30, 40, 50 or older? How long will they live? Are there things that will shorten their lifespan and what are those and how do we, you know, prevent that?
Mike: Got it okay thank you two more so the bottom three questions and I'm just I'm sorry i'm prioritizing the parents questions before I get to the other ones. So Hans says is provider hesitancy to refer NGS going away or stubbornly sticking around and then there's another question on we often (I think this is related so I'm going to bundle them) we often come across stories of clinicians telling undiagnosed families that sequencing won't change the management of the family's disease and/or insurance can't cover it which of course ridiculous because we have Behind the Seizure but that was editorial what would you say to those doctors what can we do as patient advocates to get more diagnoses and then she's got to follow on what are your thoughts on a nationwide sequencing study similar to the DDD in the UK?
Dr Mefford: That would be great. I'd love to do that. As far as I think the insurance question is a hard one and you know not changing management not, you know, I think that's not necessarily true. So, you know, when I write my notes we always talk about at least the handful of genes where we'll change management because usually we're asking for a panel or an exome and we can say you know the phenotypes overlap it cannot predict by looking at this child clinically what their mutation will be and it's important because the diagnosis will change the management and it'll also give prognosis information. If insurance still won't cover the cost... find a research study, you know, we enroll kids all the time you know there are many groups around who are studying this. Studying you know similar phenotypes and are willing to enroll kids I think you know it's always I'll be honest you know sometimes it's challenging for us too right because I'm not a diagnostic lab and so you know me finding a mutation doesn't go into your clinical you know it doesn't necessarily go into your electronic medical record. So there are challenges that way but I think but but usually you can find someone who's willing to do that. I do think that you know encouraging the other thing is that and it's a hard sell for insurance companies but you know having a diagnosis puts you in a position right to be ready when it will change management right. It puts you in a position to be ready to join the clinical trials when they're available for a new therapy that might be effective so it might not change your management today but it might change it tomorrow or next week or next month and I personally think that's important most insurance companies would say well that's that's still research so we're not paying for that but that's ridiculous. So as you know how do you advocate? I think you know working with your physician to say no it's you know in some cases it does change management and finding you know researchers and physician researchers like myself and many others that are around say to help you fight those battles it's not easy I know there's nothing more frustrating for me than when I can't... a person's insurance won't cover the testing it's like what do you do? Yeah and then Hans said the provider hasn't seemed to refer Next Generation Sequencing going away or stubbornly sticking around and does that just fall under the insurance answer? Well I think geneticists aren't hesitant to send testing you know I think that some some neurologists might be hesitant some general providers might be hesitant because the interpretation is important right and so I think you need to be able to to interpret the results of the test that you send and give that information back to your patient or engage you know the right people geneticists genetic counselors you can do that. I don't Next Generation Sequencing is not going away so I'm hopeful that, you know, our neurologists here are much much more comfortable sending those tests. They're not always still comfortable with the results but they're at least getting the testing done.
Mike: Yeah it's weird to tell parents that they're lucky when they have a diagnosis they sort of look at you like this is the worst news ever you're like well yeah there's a lot of people who never made it this far. Okay so we got four questions left and we're at the top of the hour so I'm gonna if anyone has asked a question throw them in so at the top epilepsy is a tricky phenotype to study and model what are some animal models or in vitro models that Dr. Mefford thinks are particularly helpful for studying genetic epilepsies?
Dr Mefford: Depends on the gene. You know, mouse models are generally helpful but they're expensive and they take time there are people developing zebrafish models for lots of epilepsy-related genes that are faster and cheaper and allow kind of high throughput drug screening so that's happening and cellular models. I think, you know, all these all have caveats of course but, you know, cellular models and making things like organoids from stem cells is likely to be helpful down the road as well.
Mike: Yeah we're excited about what Coba is doing there. He's building a bunch of patient derived organoids. And I'd love to hear your thoughts on how your (this is their words not mine) how her incredible analysis of 1200 DEE cohort helped our understanding of the clinical overlap in epilepsy, ASD, and other developmental disorders many thanks to Dr. Mefford for this seminar.
Dr Mefford: Yeah that's been fun we've kind of always known there was an overlap but I think that's just become more and more evident and so we're working on a paper right now where we sequenced 200 some undiagnosed kids and their parents and tried to identify the underlying cause and about half of them where we've been able to make a diagnosis it's for neurodevelopmental disorder. Not as what you think of your DEE. I think, you know, so I think there's a lot of there's a lot of overlap. I do think there are certain genes like SYNGAP1 where if you have a pathogenic variant in that gene your risk for epilepsy is high. There are others where it's not. So there are definitely some distinctions. Also a lot of blurring. I also think it's important to remember that I test patients who come from neurologists who study epilepsy. So their primary diagnosis is epilepsy. I work with other providers who see kids with neurodevelopmental disorders so they might focus on the developmental piece and say "oh, by the way they have epilepsy too". So some of it is the bias, you know, an ascertainment bias and kind of where they came from and what diagnosis they're "labeled with" which, you know, could probably be one of three different diagnoses or four right. Is it ID with epilepsy or epilepsy with ID? Is it autism with epilepsy or is it a DEE? Some of them are distinctive but many are not.
Mike: Yeah or is it just an unholy trinity? Oh so anonymous attendee says is it possible to test epilepsy and family members to predict if a future child a family will have epilepsy?
Dr Mefford: I guess the question here is inherited versus de novo and maybe yeah you know the inherited and milder epilepsies tend to be more complex so there's often not a single gene that's not true for all is it possible to predict if you know the mutation in a family member who has epilepsy then yes you can use that information to test other family members if they're interested but a lot of the times it's it's complicated and for a lot of the DEEs these aren't inherited. I think it's important to remember that they can be, right. A parent can have you know 10% of sperm or eggs that have a mutation and can have a second affected child that's an important thing to discuss with your geneticist and prenatal provider if you have one affected kid.
Mike: Right. He doesn't stop.
Dr Mefford: I can talk a lot about that paper.
Mike: Do you want to do that next?
Dr Mefford: Sure I can do that quickly. So it's exactly what I was talking about. So that we wrote a paper looking at the risk of having a second affected child when it's a new mutation in the first child and this speaks to what I was just saying what we call parental mosaicism or gonadal mosaicism as a parent. We've known for decades that this this happened in lots of different disorders where the mutation appears to be new in the child and most of the time it is and it just happens once it's a single sperm or egg that's affected it's passed on the child has the disease and it never happens again but what we found when we looked really carefully at parents who had a child with a de novo mutation causing epilepsy and now other studies have done a similar thing, five to ten percent of parents are actually mosaic for the mutation which means that it's not a zero chance that that will happen again. It's an increased chance. It's a there's a real chance that another child could be affected. It is important for prenatal counseling because we need to, you know, impart this to parents. We can't just say "it's never gonna happen again".
And yeah having one normal child does not help necessarily because if you have 10%, you can think of it... and this usually happens in the sperm, so you can think of it as 10% of the sperm were affected right because maybe there's a 10% chance that it'll happen again but a 90% chance that it won't. So what's difficult is actually if we can't find it in the blood or the saliva of parents we can't find a mutation that's reassuring but it's not 100% because it could be present in sperm. We can test sperm. Not many places do it but it can be done. We can't test eggs. It's too invasive. So yeah so I think it's important to remember that, you know, even if it doesn't look like the parents are mosaic it can happen again and if it's something you're concerned about you know there are options for prenatal testing for pre-implantation genetic diagnosis as far as de novo disorders but you know these are all things that can be discussed with a prenatal provider.
Mike: And there's one last question on here about blood brain barrier in treating epilepsy and how it affects drug treatment. I guess the question here is around delivery? Does that complicate?
Dr Mefford: It does yeah I mean you have to get it across that blood brain barrier so it does complicate things in that you have to have the right type of you know small molecule or therapy. I don't know enough about the specifics to say what works and what doesn't but I think it's it's always a consideration when you're treating a disorder like this, that you get it into the right cells.
Mike: Right awesome thank you so I have one last question just because we we mentioned that we did find that that 65 year old patient and we're so excited about it we're actually doing a short movie about it so look forward to seeing that in a few months but what was cool there what I heard about was normally when people say trio testing they test kid and mom and dad to compare but in the case of this patient, if I've got the story straight, mom and dad weren't around anymore so they tested patient and both of her sisters how does that work through like how do I explain that, like, is that harder to do? Is it...?
Dr Mefford: Yeah so that's similar to Hans' question I think and that, well, sort of, they probably did that... so for example if you have a VUS which is another thing you can do with the VUS is to say well, I've got three unaffected siblings. Let's test them. If they have that same VUS probably not causative. It's probably a benign variant, right, if they're healthy. So testing the sisters in that case helps if they're negative for the same change in saying okay they're not affected they don't have the change she's affected she has the change or he's affected he has a change. That helps. It's not 100% but it's certainly supportive evidence right and if in that individual if it was a truncating mutation, for example, you could be pretty confident and testing the siblings again helps it may actually not be necessary right because we know that truncating variants yeah but simply extended family members can help with the VUS's sometimes.
Mike: got it and it was cool about that is they they weren't testing for the they were trying to rule because because dad wasn't around they were trying to rule out okay is our current health issues inherited so let's make sure there isn't some known I don't know whatever and they stumbled across it and it just that story's so exciting because it begs this question how many more are there who we could be helping that we don't even know about?Dr Mefford: They're definitely out there for sure.
Mike: On that happy note I can't I don't have the words to thank you for everything you do and for doing this webinar. My pleasure. So we are grateful.
Dr Mefford: I hope it was helpful.
Mike: It was.
Dr Mefford: You guys are doing great work. Keep it up.
Mike: Thanks. Thank you so much. Bye-bye. Thanks. Bye.