Serotonergic Mechanisms of Seizure-Induced Central Apnea Seminar

Respiratory dysfunction following generalized convulsive seizures is proposed as a causal mechanism in SUDEP. Central chemosensitivity to carbon dioxide, or CO2, drives ventilation in response to hypercapnia, especially during sleep. Dr. Richerson and colleagues have found that this response may be impaired in people with epilepsy, and that convulsive seizures lead to prolonged inhibition of the ventilatory response to hypercapnia.


You will hear

In a mouse model of Dravet syndrome, Dr. Richerson’s team has obtained evidence that seizures inhibit serotonin neurons that detect changes in CO2 levels. Defining the mechanisms of seizure-induced central apnea may lead to novel preventative treatments for SUDEP.



About the Speaker

The seminar was presented by George Richerson, MD, PhD, Professor and DEO/Roy J Carver Chair in Neuroscience at the University of Iowa, Carver College of Medicine.

Download Full Transcript


This seminar is part of CURE Epilepsy’s Frontiers in Research Seminar Series. This program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.


Understanding Post-traumatic Epilepsy and Impacts on Quality of Life

Dr. Pugh described the association of lifetime mild traumatic brain injury (mTBI) history, epilepsy, and post-traumatic epilepsy (PTE) with quality of life outcomes in post-9/11 veterans. Dr. Pugh’s work uses survey data as well as data from the Veterans Health Administration. Worse quality of life was found for persons with PTE vs either epilepsy or mTBI alone. This research demonstrates the impact of lifetime TBI and the importance of measurement of health outcomes in epilepsy. 

MARY JO PUGHAbout the Speaker
The seminar was presented by Mary Jo Pugh, PhD, RN, a retired Air Force nurse, and Professor, Department of Medicine, Division of Epidemiology at the University of Utah School of Medicine. She is an investigator at the Informatics, Decision-Enhancement, and Analytic Sciences Center of Innovation (IDEAS COIN) and the VA Salt Lake City where she leads the COIN focus area that aims to identify and mitigate health risks in Veterans across periods of vulnerability. Over the past decade, she has targeted her work using VA data to identify phenotypes in populations with complex comorbidities such as those with traumatic brain injury (TBI) and complex multi-symptom illness in Gulf War-era Veterans. Dr. Pugh’s current work related to TBI includes longitudinal observational and prospective studies to identify the emergence of distinct neurodegenerative conditions such as cognitive impairment and epilepsy and clusters of multimorbidity that may have similar or networked biological underpinnings.

This seminar is part of CURE Epilepsy’s Frontiers in Research Seminar Series. This program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.


Understanding the Mechanisms of Epilepsy in mTORopathies (FCDII and TSC)

Hyperactivation of mechanistic target of rapamycin (mTOR) signaling is implicated in a number of focal cortical malformations associated with intractable epilepsy. While the link between focal cortical malformations and epilepsy is well known, the underlying mechanisms remain unclear.

Dr. Bordey focuses on the role of hyperpolarization-activated cyclic nucleotide-gated potassium channel isoform 4 (HCN4) as an mTOR-dependent driver of epilepsy in tuberous sclerosis complex (TSC) and focal cortical dysplasia II (FCDII).

Download Full Transcript

About the Speaker
The seminar will be presented by Angelique Bordey, PhD, Professor of Neurosurgery and Cellular and Molecular Physiology at Yale School of Medicine. She is member of the CURE Epilepsy Scientific Advisory Council and a former grantee.

This seminar is part of CURE Epilepsy’s Frontiers in Research Seminar Series. This program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.


Q&A with Dr. Angelique Bordey

The resting membrane potential in the Rheb neurons, was it normalized during your current-injection test?

Dr. Bordey: Whether it was holding the cells at the resting membrane potential, which we did to inject. We held them at the resting potential and then injected current-injection. Like it would be more physiological condition. I’m not too sure it was a proper answer to the question, but I see it. They are depolarized compared to controls.

Do you see HCN4 staining in all FCD-two patient samples or only in a subset of samples?

Dr. Bordey: So far we saw in all of them. Yes, in all of them. We do not know the gene variance for these FCDs. As I know there are 30% or more FCDs that are due to mTOR gene variants, I don’t know which one we had. There were classic or FCD-two and I think we reported it in the paper. Some of them had a balloon cell and some did not. So type A and B.

Do you see a translational path, and I had the same question, do you see a translational path to treat people?

Dr. Bordey: The answer is yes. So the beauty of the HCN4 is that it’s not in the cortex, at least in adults or young adults. So it is a perfect target for gene therapy. It is expressed during development, it has not been studied very much, but it is there, and after birth it decreases tremendously. It’s expressed only in the thalamus, I think the amygdala and the cerebellum. So if targeting the cortex is a good gene therapy target, then can we use a drug? Presumably not because that would slow down the heart, the HCN4 is expressed in the heart and is important for their pacemaking activity, which I did not say, there are pacemaker channels, which also explain why they would maintain seizures, activity or cell firing.

Can we consider antisense oligonucleotide? Possibly. I think there is a lot of development by companies with small SIRNA that can be delivered intrathecally and they will last for about six months. HCN4 may be a good target. I think we need to know what is the function of HCN4 in the thalamus, because we would not want to block it there. It may be in the hippocampus a little bit. So it would be worth trying. I think it’s important actually to try whether a systemic SIRNA injection for covering the brain would work and have no side effects. So this is a big question. It’s just easier than gene therapy to move towards clinical application. So those are the two alternatives I can think of.

Have you tried blocking HCN4 after seizure onset?

Dr. Bordey: We did not. This is a very good question. We wanted to do it. We have the plasmid, we have not done it. We know you can block… When seizures are established, we know that we can block them and it is sufficient to… We can inject drug and it’s sufficient to block seizures. We have done that with Lena who showed that blocking translation after the onset of seizures, after they establish well decreased seizure frequency. So you can shrink the cell size and presumably remove ion channels since HCN4 is translation dependent. We have also shown that if you block a molecule called [philomena 00:29:54], and we published that last year. If you block the activity of philomena with a small molecule, once the seizures are established, so in mice that are, I think there were maybe two months of age, you do decrease seizure frequency. So I’m hopeful that doing that also in adult will work.

Do these genetic epilepsies respond well to vagal nerve stimulation or resective surgeries?

Dr. Bordey: Yes, resective surgeries. Yes. They respond well. I think they’re… I don’t know all the numbers on top of my head, but maybe 20 or 25% of the patients will go through surgeries, I may be wrong with the number, but roughly. Not everybody can go through surgeries, depending on the location and if they have too many of these malformations. Patients can have 1 to 50, which makes things complicated. And I think at least 50% of the patients will be seizure-free, I think. I think it depends on the clinical center, maybe 50 to 70, but very often seizures come back. So that’s a problem with the surgeries. The vagal nerve stimulation, I know it is used, and I do not remember the numbers, the test statistics in terms of efficacy. So I don’t know that.

Do you know the upstream genetic drivers of increased HCN4?

Dr. Bordey: We know it’s rapamycin sensitive, we saw it. So we use the plasmid Rheb, and Rheb is right upstream mTOR, and this is sufficient to increase HCN4. We have, I don’t think we published it, but we had a [inaudible 00:31:57] in TSC one [flocks 00:31:59] mice, so mute knockout mice. In TSC one, HCN4 four was increased as well. And since it’s increased in the patient, presumably it includes other genes since TSC one and Rheb, and then upstream, you have AKT and PI3K. But we have not looked at PI3K, we have not looked at P10 [inaudible 00:32:23] five, for example, to get to one complex, which I think is really important to do.

Do you know if it’s translational or is it the transcription? The up regulation?

Dr. Bordey: So it’s for ADP dependent, like I think Lena had shown, so it’s translation dependent. There is no increase of the MRNA, at least what we saw in our mice. And it’s not unexpected that you have cells with floating MRNA that is not translated, so we clearly promote a translation of that gene.

Is the HCN4 MRNA the same isoform as found in other parts of the brain or the heart? And if not, there may be potential for targeting?

Dr. Bordey: Yes, this is a good question. And I’ve asked myself that question and we have not, we have not checked. It is a possibility. So there are two different isoforms, a long and a short isoform, and I don’t remember which one the brain has, but the problem is we have not looked in the disease condition, which could even be a slightly different isoform. So I think that this is a very… This is something really important to do.

Do you know whether HCN4 is associated with other types of epilepsies?

Dr. Bordey: Yeah, actually I don’t remember the group that did it, but they found… I actually don’t remember right now, I should have checked. I know there is a mutation in HCN4 channel that has been reported, I think maybe last year, that is associated with seizures. I don’t remember if it’s a gain of function or loss of function. So there’s one study that reported that. There’s also genomic association studies showing that it actually could be associated with also a bipolar disorder since you control excitability. So I think this is quite… This is a gene of interest that people have not seen in the past because it’s a very discrete expression in the brain, so people thought very limited function and you have no clean blocker of HCN4. The blockers will touch all the HCN channels, so it has been hard to study by many. But now I think that our study highlights the importance of this channel in seizure, and perhaps, other disorders.


Postdoctoral Data Blitz on Translational Research in Epilepsy

CURE Epilepsy’s Frontiers in Research Seminar Series program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.

CURE Epilepsy conducted a virtual, postdoctoral data blitz, a series of short talks, to provide senior postdoctoral researchers an opportunity to share their work with other epilepsy researchers. The seminar focused on highly innovative, translational epilepsy research that can lead to a cure. The hour-long seminar highlighted work from the following three postdoctoral researchers, who each had ten minutes to present their work followed by five minutes for questions from the audience:

Maria Belen Perez-Ramirez, PhD
Stanford University
Focal Status Epilepticus and Gabapentin Effects on the Neocortex

Abstract: Status epilepticus (SE) is a severe neurological disorder with consequences ranging from death to increased risk for subsequent unprovoked seizures leading to epilepsy and cognitive abnormalities. An important unanswered question is whether a prolonged episode of focal status epilepticus (fSE) in the neocortex elicits long-lasting modifications leading to epileptogenesis. Dr. Perez-Ramirez is testing the hypothesis that alterations after a single episode of fSE result in structural and functional changes among neurons in neocortical circuits, potentially leading to epileptogenesis, and that treatment with an anti-synaptogenic agent, Gabapentin, may have prophylactic effects on structural, functional, and behavioral abnormalities following fSE.

Maria-Belen Perez-Ramirez is a biologist and obtained her PhD degree in Biomedical Sciences at the National University of Mexico. She is currently working as a PhD fellow in the neurology department at Stanford.

Lena Nguyen, PhD
Yale University
Targeting Translational Dysregulation in mTOR-related Epilepsy

Abstract: Hyperactivation of mTOR signaling due to mutations in mTOR pathway genes causes a spectrum of neurodevelopmental disorders (mTORopathies) associated with malformation of cortical development and intractable epilepsy. Dr. Nguyen’s research has led to the discovery that altered activity of the mTOR-dependent translational repressor 4E-BP1/2 contributes to neurodevelopmental defects and epilepsy in two prototypic mTORopathies, tuberous sclerosis complex (TSC) and focal cortical dysplasia type II (FCDII). The team is currently investigating novel and clinically relevant strategies (such as gene therapy) to target 4E-BP1/2 to treat epilepsy.

Lena Nguyen is a postdoctoral associate in Dr. Angelique Bordey’s lab at Yale University School of Medicine. Her research focuses on how alterations in intracellular signaling pathways contribute to epilepsy in neurodevelopmental disorders.

Rajeswari Banerji, PhD
University of Colorado
Enhancing Glucose Metabolism via Gluconeogenesis with PK11195 is Therapeutic in a Zebrafish Model of Dravet Syndrome

Abstract: Energy producing pathways are novel therapeutic targets for the treatment of neurodevelopmental disorders. Dr. Banerji’s research utilized a translatable zebrafish model of Dravet syndrome (scn1lab) which exhibits key characteristics of Dravet syndrome and shows metabolic deficits accompanied by downregulation of gluconeogenesis genes, pck1 and pck2. Treatment with a pck1 activator normalized dysregulated glucose levels, metabolic deficits, translocator protein expression and significantly decreased electrographic seizures in mutant larvae. Inhibition of pck1 in wild-type larvae mimicked metabolic and behavior defects observed in scn1lab mutants. Together, this suggests correcting dysregulated metabolic pathways can be therapeutic in neurodevelopmental disorders such as Dravet syndrome arising from ion channel dysfunction.

Rajeswari Banerji is a senior postdoctoral trainee in the Department of Pharmaceutical Sciences, University of Colorado, Anschutz Medical Campus. Her research involves understanding and developing novel metabolism-based therapies for neurological disorders, particularly epilepsy.

Download Full Transcript


Let’s start with a question that came in for Lena. They want to know, why did you focus on 4E-BP1 and not 4E-BP2?

Dr. Lena Nguyen: That’s a great question. There are three sub-types of 4E-BP, there’s 4E-BP1, 2, and 3. 1 and 2 is expressed in the brain, with 2 being the most highly expressed. At the time that we started this study, the tools for targeting 4E-BP1 was more readily available to us since this is the most common studied subtype in the cancer field. That is why we used it. Functionally, 4E-BP1 and 2 are very similar. The sequence hemology, I think is more than 60%. In our case, we over-expressed 4E-BP1 which would compensate for both the inhibited activity of 4E-BP1 and 2. I think that expressing constitutive active 4E-BP2 could potentially have the same effect.

Here’s a question for Belen. They want to know, how do you analyze your EEGs? And what criteria need to be met to establish that there is seizure activity in your models?

Dr. Maria Belen Perez-Ramirez: Thanks for the interest. To analyze the EEG, we use Power Analysis. Please remember that we induce the seizure activity at the beginning but we haven’t still studied later EEG activity. But then to establish that there is seizure activity, there needs to be interictal electrographic discharges, and then I also expect to have behavioral contralateral jerkings from the animal. When the animal has these contralateral behavioral seizures together with the interictal seizures, then that’s when I start counting the time, the two hour period of seizure activity. Then something we notice is that after their interictal spikes, they can develop to ictal activity. Then we just let the seizure activity for two hours and then we stop the seizure. But there should be both behavioral and electrographic interictal spikes.

I’m going to move onto Raji. There’s a question here, and they want to know please tell us more about PK11195? Do we know anything about the structure? Does it get into the brain? Has it been tested in other models of epilepsy?

Dr. Rajeswari Banerji: That’s a great question. Yes, we know a lot about this drug. This had been discovered in 1977, so a lot of research has been done and so this drug is a synthetic GSP11, so it binds to the mitochondrial ligand that is the translocator protein. It is also known as the peripheral benzodiazepine receptor, so it does have a lot of information on how it mines and how it producing neuro [inaudible 00:41:38], so there’s a lot of research going on. As I mentioned in the talk, there are a few papers which have shown an anti-epileptic property of this drug. We are really excited to try this in our new model, and also try it in mice models.

Raji, do you know if the efficacy of this drug, this anti-seizure activity, is that dependent on its improvement of the metabolic activity or are they two separate things?

Dr. Banerji: No, what we think, certainly after doing three years of research on our zebrafish model, that a metabolic deficit is required for this drug to work. We think this drug may work as a censor, actually, because we don’t think it always improves. It may reduce the glucose levels, also. A better term will be normalize, so it senses any kind of defects, metabolic defects, and then corrects it. We still don’t know how it corrects a seizure phenotype, and that’s something which we have to understand using other models maybe. But that’s a good question also. Thank you.

Lena, back to you. The high variability in your seizure rescue finding is really interesting. Do you have any clues on why many saw complete reduction in seizures, while others looked similar to the control group? Did you find that seizure frequency correlated with any of your other measures?

Dr. Nguyen: That’s a great question. Several things could contribute to this, one of which is a technical consideration for in-utero electroporation. We have been pretty good at getting consistent with the area that we target and the size of the targeted area, but when electroporating multiple plasmid as we did, not all cells are going to end up with all of the four plasmid. The majority will, but not all. These differences could contribute to the large variability. In addition to that, differences with combination efficiencies could lead to different levels of 4E-BP expression, which could also contribute to this. In our case, cell size is a good readout of whether 4E-BP activity worked the way we had intended it to, and so we have correlated cell size with seizure activity and we found a very strong correlation where those with no or low seizure activity following 4E-BP expression had smaller cell size, and those with higher seizure frequency had larger cells, suggesting that 4E-BP might be insufficient. It is interesting that we still see such a strong effect despite this variability, and we might actually be underestimating the effects of 4E-BP on seizures.

Lena, have you looked at or has anyone else looked at different models of cortical malformations to know is there 4E-BP1, is it also hyperphosphorylated in other models? Is that known?

Dr. Nguyen: Other groups have shown that [inaudible 00:45:23] is increased in human mTORopathies, like TFC and FCD2, and increased phosphorylation is expected as well in other mouse models.

Next question is back to Belen. What would your next steps be to translate your findings to people who have experienced status? Have you tried Gabapentin animal models of epilepsy?

Dr. Perez-Ramirez: Thanks. Yeah, that’s a complex question, the first one. Yes. A bit, first, I think we need to find what will be the best regime of Gabapentin, or Gabapentin treatment. Meaning the concentration and timing, because also there is a time window as far as we know where this is synaptogenic activity. We still don’t know in this model, what is this time window and the latency? That’s also another issue we need to address. Also, it will be important that once we know the most efficient Gabapentin treatment regime, then how this translates to human. Gabapentin is already an FDA-approved drug, so we just need to understand how the doses even after status epilepticus can work in humans. Then as I was saying, the window will be very important as also Gabapentin, it’s as far as we know, the efficacy happens also with newly formed synapses. That’s why it is important to know the latency, or the synaptogenic period.

About the second question, was Gabapentin trialed in other models? In the lab here, it was already tested on the undercut model of epilepsy, and they found similar results with decreases in alpha 2 delta 1, and decreases in synaptic activity, and also decreases in synaptic connectivity. Then there is the group from Rossi, and they tried Gabapentin in the pilocarpine model of temporal epilepsy, and I think they found also that Gabapentin was able to decrease astrocytic activity. In the latest article, they test this with animals that were treated with PILO and then treated with Gabapentin, and then given a second dose of PILO, and they found that there was decreased mortality, decreased racine scale achieved, and decreased animals that get to actual status epilepticus on that second dose. And then, there is the group of [inaudible 00:49:01] that they tried Gabapentin on a model of cortical malformation, and they also found decreased epileptiform activity with Gabapentin and I think also decreased expression of thrombospondins. It’s been tried in different models.

You mentioned about timing, so have you looked at, I think you’re doing it at three hours after status, the Gabapentin treatment you initiated? Have you tried other time points, like starting it later?

Dr. Perez-Ramirez: That sort of thing, we are interested in looking at.

Thank you. I had another question for Lena, just a technical question. In the samples from the humans where you showed the 4E-BP1 hyper-phosphorylation, what did you use as controls for those?

Dr. Nguyen: Because they were human samples, [inaudible 00:50:20] come across, we have not been able to get control human tissue samples. But there have been other studies where 4E-BP at this level is not found in the control tissue. We also looked at surrounding cells just to see. It’s not the best control, but those cells did not have increase. It was very specific to the dysmorphic enlarged neurons, the increase.

Abstracts Due: Postdoctoral Data Blitz on Translational Research in Epilepsy

CURE Epilepsy’s Frontiers in Research Seminar Series program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.

CURE Epilepsy will conduct a virtual, postdoctoral data blitz, a series of short talks, to provide senior postdoctoral researchers an opportunity to share their work with other epilepsy researchers. The seminar will focus on highly innovative, translational epilepsy research that can lead to a cure. The hour-long seminar will highlight work from three researchers, who will each get ten minutes to present their work followed by five minutes for questions from the audience.

More information on the researchers and their talks will be available in early April.

Interested speakers may submit an abstract accompanied by a biosketch to for consideration by Friday, February 26 at 9 PM EST 2021.

To be eligible, you must

  1. Have at least three years of postdoctoral experience
  2. Be working on a novel, translational research project on any of CURE Epilepsy’s priority research areas which include acquired epilepsy, drug-resistant epilepsy, pediatric epilepsies, SUDEP, and sleep and epilepsy.

Abstracts must be no longer than a half page (250 words) in length and single-spaced. Figures and/or graphs are not required. Please include your name, email and institution, and any publications related to the work in your abstract.

Abstract submission deadline is Friday, Feb. 26 at 9 PM EST. Email your abstract to You may also reach out to us at this address with any questions regarding this seminar.

Researchers who are selected to speak will be contacted via email by end of March 2021.

The Use of Organoids in Epilepsy Research

Human brain organoids derived from human pluripotent stem cells are a powerful testing platform to model and study epilepsy. These organoids may be advantageous over traditional rodent models which do not always exhibit human pathology, possibly due to differences in size, complexity, and gene expression patterns between rodent and human brains. Due to their greater structural complexity and a diverse population of neurons, three-dimensional brain organoids are a useful tool to model epilepsy and test potential therapies.

Dr. Parent speaks about recent advances in techniques using brain organoids and their use in epilepsy research.

Download Full Transcript

This seminar is part of CURE Epilepsy’s Frontiers in Research Seminar Series. This program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.

About the Speaker
The seminar will be presented by former CURE Epilepsy grantee Jack Parent, MD, William J Herdman Professor of Neurology and Co-Director of the Epilepsy Program at the University of Michigan.




Q&A with Dr. Jack Parent

Did you see evidence of variable or incomplete X inactivation in the female organoids?

Dr. Parent: That’s a really good question. When we started the PCDH19 studies, we started with female patient-derived IPS cells. We were doing 2D cultures. And we saw evidence for X inactivation, which we wanted. But in any given culture, we didn’t know what percentage were expressing the mutant and what percentage of cells were expressing the wildtype, and it would be just too much work to try to figure that out for every experiment. I think one of the reasons they’re asking the question is with prolonged passaging, you can get erosion of X inactivation for some genes in human pluripotent stem cells. So because of that problem, we changed approaches and we took complete knockout and wildtype and mixed the two in order to make a mosaic model where we didn’t have to worry about escape from X inactivation. That’s an important point.

Have you used electrophysiology to assess abnormal activity in the STRADA organoids, and if so, did you use any mTOR inhibitor to rescue it?

Dr. Parent: Both very good questions. Louis has done some experiments using rapamycin to rescue of the morphological phenotype. It rescues the megalencaphaly in the budding phenotype. In terms of the physiology, we published some work with 2D cultures looking at the physiology of the STRADA organoids. We see some changes, but it’s minor. And those are with dual SMAD differentiation of excitatory cortical neurons. But we think the best approach is going to be combining excitatory and inhibitory fusion organoids. We have not done that yet in STRADA, and that’s one of the things we want to do, looking at MEA recordings, but also putting in the patch electrodes to record local field potentials like the UCLA group is doing. I actually went on sabbatical there to learn it.

In your PMSE organoid model, you talked about how you see these different phenotypes at different ages at different time points for the organoids. Do you have an idea of how that might correspond to the human brain development age?

Dr. Parent: That’s a good question. At the time points we’re looking, we probably never got later than mid-second trimester stages. To get to third trimester and even postnatal stages, you really have to go nine months a year. It really is like the time course of human development. So you really have to culture them very long. It makes for very long experiments and unhappy grad students and postdocs. People have cultured them for four years, even longer.

The information contained herein is provided for general information only and does not offer medical advice or recommendations. Individuals should not rely on this information as a substitute for consultations with qualified health care professionals who are familiar with individual medical conditions and needs. CURE strongly recommends that care and treatment decisions related to epilepsy and any other medical condition be made in consultation with a patient’s physician or other qualified health care professionals who are familiar with the individual’s specific health situation.

The Role of Adenosine in Epilepsy

Adenosine is a well-characterized endogenous, anticonvulsant, and seizure terminator in the brain. Overexpression of adenosine kinase, the major adenosine metabolizing enzyme, directly contributes to seizure generation due to depletion of extracellular adenosine. Conversely, therapeutic levels of adenosine can suppress seizures in rodent models of temporal lobe epilepsy. In addition to the anti-seizure effect of adenosine mediated via the adenosine receptor, adenosine kinase may also have disease-modifying effects mediated through regulation of DNA methylation, making it an attractive therapeutic target.

Dr. Boison talks about the anti-seizure and disease-modifying effects of adenosine in epilepsy, as well as recent advances in developing adenosine kinase inhibitors as therapeutics for epilepsy.

This seminar is part of CURE Epilepsy’s Frontiers in Research Seminar Series. This program is generously supported by the Nussenbaum-Vogelstein Family and aims to help educate and expose researchers, clinicians, and students to exciting epilepsy research and also provide opportunities for young investigators to interact with leaders in the field.

Download Full Transcript

About the Speaker
The seminar is presented by Catalyst Award grantee Detlev Boison, PhD, professor at the Department of Neurosurgery at Rutgers University.




Q&A with Dr. Detlev Boison

Have you observed or looked into DNA methylation differences in response to adenosine?

Dr. Boison: That’s a good question. We have primarily seen changes in neurons. There are new data which I haven’t shown today, but we find the trends in ectopic expression of ADK-L in neurons during epileptogenesis. This is really a transient period and might be the reason why trends in treatment with an ADK inhibitor perfectly tailored to this trends in over-expression of ADK-L in neurons, which really coincides with increased 5-mC during the same time span might be an important role for the epileptogenic process.

Over time, do the medications change the seizure onset or property of the seizures?

Dr. Boison: The goal here is basically to have a prophylactic treatment to initiate, and yet to have trends in treatment during the latent period of epileptogenesis with the ultimate goal to prevent epilepsy all together. Now in most studies, which I’d shown previously, we compared the seizure phenotype at six weeks and nine weeks after initiating epilepsy and we did not find any differences. Of course, it would be worthwhile in future studies to extend that time period to longer time spent looking at brains at three months or six months and we will certainly do this in the future.

Are there any commercial assays available for measuring adenosine levels in human plasma?

Dr. Boison: No, there are not. Adenosine can be quantified by HPLC and by LCMS/MS methods. There are no commercial assays. The blood-brain barrier is not really penetrable for adenosine and adenosine in the circulation has an extremely sharp half-life in the range of seconds. So plasma adenosine is unlikely to be a representative for adenosine levels in the brain.

How is astrogliosis linked with epilepsy? Do we know if it causes epilepsy or is it just correlated with epileptic seizures?

Dr. Boison: There’s a lot of data, not just from my group, but also from other researchers that suggests that astrogliosis can truly be a cost for the generation of epileptic seizures. And astrogliosis is triggered by all the inflammatory processes that trigger the initial phases of epileptogenesis. So you get microglial activation and astroglial activation as a response basically to brain inflammation.

How specific is 5-ITU for ADK?

Dr. Boison: It’s one of the old inhibitors. It’s not selected for ADK-L versus ADK-S. According to our findings, it blocks both. That’s the reason why it prevents epileptogenesis, but it also has sedative side effects because it also increases extracellular adenosine.

Do you know of a mechanism by which adenosine may build up in the nucleus and make methylation?

Dr. Boison: The deficiency of adenosine in the nucleus would drive DNA methylation. If there’s a mechanism that increases adenosine in the nucleus, then it would block DNA methylation and this could be directly linked to factors that inhibit adenosine kinase in the nucleus. But this is currently unknown.

How did you go about designing a biopolymer that could release adenosine?

Dr. Boison: Whenever you have a release system, you get a logarithmic release profile. If you, let’s say, just encapsulate adenosine in the membrane, you will have a burst release: Most of it will come out immediately and then drop off. So in order to get a stable release, you basically need to combine several logarithmic release profiles in one implant.

So the way those silk polymers were engineered was: First we encapsulated adenosine as micro-vesicles in a silk membrane, and then those micro-vesicles were embedded in a 3D silk metrics. Then the whole construct was coated with alternative layers of adenosine and silk, always repeating, and then you can cap the whole thing with several additional layers of silk to slow down the release. So if you do it right, then you get the release profile.

The beauty of silk is it’s a biopolymer. It doesn’t have any toxins. It’s fully resolvable. Silk has been used as a suture material for decades and it’s a relatively boring amino acid repeat of glycine and alanine. So it’s very biocompatible.

What are good targets for DNA methylation to measure?

Dr. Boison: We are primarily interested just in the global DNA methylome, because based on the evolutionary principles, I think this is a primordial mechanism to regulate the entire DNA methylome. Because if you think in terms of evolution, if you want to regulate genes, you cannot start with transcription factors because transcription factors need their own genes, they need to be controlled by protein kinase pathway, which all needs their own genes which need to be regulated. This is way too complicated.

Now, if you really want to develop a regulatory system for the genome, the most simple way you can think of is just adding and removing methyl groups. This affects the entire genome and I believe this was probably one of the first mechanisms that came up in evolution to regulate the entire DNA methylome on a global level, and everything else that creates target specificity was invented later during the evolution.

There is a question about your MRS drug and whether it is BBB permeant: Does it permeate the blood-brain barrier?

Dr. Boison: We have evidence that it goes through the blood-brain barrier. We’ve also engineered ADK-L mice, which over-express ADK-L in the brain and we can find effects of MRS4203 in those ADK-L mice which is a direct proof that it has an effect in the brain.

Is there any application of your research yet for treating patients who have intractable epilepsy?

Dr. Boison: The primary goal right now with those new compounds is epilepsy prevention, to prevent them from becoming intractable. To treat intractable patients with adenosine is also doable. We have shown that adenosine can prevent pharmacoresistant seizures in our epilepsy models. The challenges are for intractable patients we need long-term therapies over months or years, which means the drugs we have to use need to be very, very safe and we need to come up with solutions to avoid side-effects based on extracellular adenosine. One way to achieve those goals would be focal therapies. One approach could be gene therapy to knock down adenosine kinase in an epileptogenic brain area, but that’s not all that quick solution most likely.

What is the role of AMP in epileptogenesis? Are some of the effects of adenosine kinase due to increase of AMP?

Dr. Boison: Most likely not because AMP levels in cells are 100,000 foot higher than adenosine. So if you change adenosine kinase, you have dramatic effects on the levels of adenosine without changing the levels of AMP significantly.

While A1R seem to have the dominant and inhibitory effect, what roles do other adenosine receptor subtypes have on seizures, particularly in the cortex, as opposed to the hippocampus?

Dr. Boison: We need to realize that the adenosine system is highly compartmentalized. There is one kind of global tissue tone of adenosine, or which also can be considered as an extracellular compartment, which is directly under the control of adenosine kinase expressed in astrocytes and which provides tonic inhibition through activation of the A1 receptors. This is just a global tissue tone of adenosine that primarily activates the A1 receptor to provide the tonic inhibition in the brain.

But on top of that, there is also a synaptic pool of adenosine. So neurons under high-frequency stimulation can directly release adenosine, and neurons can release ATP, which is rapidly broken down to adenosine. So there is a different source of adenosine at the synaptic level. And the A to A receptor has primarily a role to find you the activity of adenosine on the synaptic level. The rationale of this mechanism is basically the forearm. So if you have a globally inhibited networks through the A1 receptor, and then you have once synapse where you have adenosine providing activation of A to A receptors, you can improve the signal to noise ratio. You can even more specific signal on the synaptic level in a globally inhibited network.

The information contained herein is provided for general information only and does not offer medical advice or recommendations. Individuals should not rely on this information as a substitute for consultations with qualified health care professionals who are familiar with individual medical conditions and needs. CURE strongly recommends that care and treatment decisions related to epilepsy and any other medical condition be made in consultation with a patient’s physician or other qualified health care professionals who are familiar with the individual’s specific health situation.

Targeted Augmentation of Nuclear Gene Output (TANGO): A novel therapeutic approach to treat SCN1A-linked Dravet Syndrome

Background: Targeted augmentation of nuclear gene output (TANGO) is an antisense oligonucleotide (ASO) technology being developed by Stoke therapeutics for the treatment of severe genetic diseases. This ASO therapy targets naturally occurring, non-productive RNA splicing events to restore normal levels of the target protein. In collaboration with Stoke Therapeutics, Dr. Isom has tested this technology in a mouse model of Dravet syndrome.

You Will Hear: Dr. Isom provides an overview of TANGO and presents results from testing ASOs in a mouse model of Dravet syndrome.

Speaker Bio: This webinar was presented by Dr. Lori Isom, the Maurice H. Seevers Professor and Chair of the Department of Pharmacology, Professor of Molecular and Integrative Physiology, and Professor of Neurology at the University of Michigan Medical School.

Download Full Transcript

Audience Q&A with Dr. Isom

Is anything known about the mutation susceptibility of SCN1A?

Oh, you mean how susceptible the gene is to mutation? Very, very, very susceptible. It’s a huge hit. It’s a huge target for hits. We knew that back in the day when I was in Bill Catterall’s lab and we were purifying and cloning sodium channels and cloning the SCN1A cDNA was just nearly impossible. Everything you did made it mutate. When we saw this later on, we were not surprised at all.

Where are you measuring expression levels in whole brain tissue or are you looking at specific brain regions?

What we do is we take the whole brain and then take a slice that goes through the cortex and the hippocampus and use that to make the mRNA and protein.

Are you able to targets cell specific populations with ASO22? For example, inhibitory cells versus pyramidal neurons to see whether the therapeutic effect may increase?

This technique, as it stands, is general, so it does not discriminate between cell type. In fact, it will hit every kind of cell in the brain, including non neurons, okay? It seems to have this remarkable effect in spite of that. Now there are and encoded another biotech firm who was looking at this very similar technique but using viruses that target to PV positive interneurons. You may have seen their work at the AAS meeting last year. They have a very similar result with the SUDEP model that we have of stopping, of ceasing, of preventing the seizures and SUDEP.

There’s a question about quantification of the protein. Are you using Western blots to do that or are you doing something else?

Yeah, we’re using the Western blots. It’s more of a high throughput, so we’ve done it both ways. When you look in the paper, we’ve shown it by Western, which is more low throughput, but for all those millions of samples, we used a more high throughput protein, a high throughput modified method.

There’s a couple questions that came in about age of administration, especially since you’re talking about P2 versus P14. How did that translate to humans?

That’s a really good question. This is where mouse models have limitations. If you ask 10 developmental biologists, what a P2 mouse translates to in terms of a human, you will probably get 10 different answers whether or not that is a newborn, whether or not that’s embryonic or not, and whether or not P21, at weaning, is the equivalent of one year. I’ve read some papers where that’s more the equivalent of an older individual. It’s really difficult to predict what the mouse ages are going to translate to with humans. That’s why we wanted to push this up to the time of disease onset to see how much we could prevent. One of the limitations of this model is that there’s such a small window between the time of disease onset and the majority of the SUDEP, so we start losing animals right around P22 to P23. There’s very little window for us to do the injection and then have the full analysis before the animals start dying. I think that’s where we’ve pushed this particular mouse model to the limit. That’s why we’ve now doing the ultimate experiment, which is the clinical trial.

Have you looked at cognitive deficits in your mouse model?

We haven’t, and that’s a really good question. I get this question all the time. That’s going to be a very, very large study. At present, we’re looking at more detail at the electrophysiological details. We’re going in and looking at sodium current and firing of the inhibitory neurons and whether or not we can reverse the disinhibition that we see. Once we get that nailed down, then perhaps we could go on and do some of the behavioral deficits as well.

You did mention that no other sodium channel genes are affected by ASO22 treatment, but did you look at the entire transcript on?

We didn’t, and that’s a good question. No, but that’s something that we ought to do. Yes.

Do you think that you might need multiple injections to maintain the effect or is the single injection going to be enough?

That’s a really good question that I get asked all the time. Based on our observation that we had a single injection at postnatal day two, and we ended our experiment 90 days later, and we still did not, we saw a single seizure and a SUDEP. The question is there may be two possibilities. Well, maybe three. Okay, so two possibilities that I can think of is that in a mouse, at least, the ASO gets the brain past a critical period of development with the increase in Nav1.1, and then after that, the brain takes over and plasticity happens and that it’s normal development ensues, or this is a really long lasting ASO. We know that, that this chemistry really protects the ASO from degradation, so when you look all the way out at P90, you can see the ASO is still there. There’s still some there. It may be that you need very little over a long time, and it’s, it’s very effective, okay, or maybe mice are not small humans, and there’s a different process there, but at least in our hands, all you need is one.

Now, you notice in the clinical trial, they have two phases, right. They’re doing a single dose, and then they’re doing multiple doses to see what effect that is. If you think about the original ASOs news [inaudible 00:55:04], right, for spinal muscular atrophy, you have to do multiple doses. I think they inject quarterly, I think it is, in intrathecally in order to keep that therapeutic benefits, so that may be true.

Have the preclinical tests been performed in other mouse models? If yes, which ones?

They haven’t. I think it would be an interesting experiment to do the test in one, in a model that actually makes a protein, okay? The models, most of the models that are out there now are haplo-insufficient. This was a null, but you could also do it in a, Dr. Yamakawa has a very interesting mouse model that has a knock-in of stop code on, okay, which is a humans variant. That would be interesting to do. Now, that’s also haplo-insufficient, so we’d expect to have the same result as we have here. It would be interesting to prove the, what I proposed that if you use a variant that actually makes a protein, that it also increases that protein and can have some deleterious effects. We should do that, but the only mouse model that we’ve used so far is the null.

What proportion of patients with Dravet would you say are good candidates for this therapy?

I think at least 60%. More than half of Dravet patients have variants that cause haplo-insufficiency. And so the nice thing about this ASO is it’s a very antagnostic, right? Because so all of those variants cause nonsense-mediated decay of that other allele and this ASL upregulates the wild type allele, which is wonderful. I think at least 60% of the, and maybe more of the, maybe as close as 80% of the Dravet patients may be helped by this therapeutic.

What is known about the location and type of SCN1A mutation related to the efficacy of TANGO in increasing protein production?

Totally agnostic. It doesn’t matter. If the mutation or the variant causes a premature stop or a deletion and targets that mRNA for nonsense-mediated decay, then this ASO will work. It doesn’t matter where. On that map that I showed you that you could have a mutation at the end terminus or the C terminus and still have Dravet due to haplo-insufficiency, the ASL will work with all of those, as long as it causes haplo-insufficiency. That’s the beauty of this treatment that it doesn’t matter where the variant is.

Biological Mechanisms of SUDEP

Background:  Sudden unexpected death in epilepsy (SUDEP) is the most important epilepsy-related cause of death, occurring in at least 1:1000 people with epilepsy each year. The risk of SUDEP increases dramatically in uncontrolled epilepsy. The events leading to SUDEP are thought to be caused by a destabilization of autonomic cardiorespiratory compensatory processes. Dr. Simeone’s research has focused on determining progressive changes in cardiorespiratory function that could increase the probability of SUDEP in preclinical animal models. Identification of temporal biomarkers that can be monitored could lead to opportunities for intervention to postpone or prevent SUDEP.

You Will Learn: Dr. Simeone discusses the progressive cardiorespiratory dysfunction seen in the Kv1.1 knockout mouse model of SUDEP and the potential role of orexin as a central regulator of SUDEP.

Speaker Bio: This webinar was presented by Dr. Kristina Simeone, Associate Professor and Director of Master’s in Neuroscience Program at Creighton University’s School of Medicine.

Download Full Transcript

Audience Q&A with Dr. Simeone

Did you notice in your knockout mice, were the animals having more frequent seizures during the sleep period as they neared SUDEP, or was that not related at all, the sleeping seizures?

So, that is a fantastic question. That is a study that we haven’t done yet. So, the study that looked at the hypnograms, that looked at sleep architecture, we analyzed the data. And then after the study was done, we sacrificed the animals for histological studies. So, we didn’t let them live until they died naturally. That study was actually conducted before we started doing that with our endpoints.

So, we know that the animals that had very disrupted sleep that they were a higher risk. They’re I think probably SD70, at the age of SD70, where 70% of the knockout colony had passed away or died of sudden death. So, I think that would be a great study. Are there more seizures that are coming out of sleep? Is the sleep architecture disrupted more as within subject as they approach sudden death? We haven’t done that. That is a fantastic question.

How do you actually calculate disruption in sleep when how much sleep does a wild type mouse get and then they’re going through these bouts of sleeping, waking, the much minimum amount of disruption that’s required to increase to SUDEP risk?

So that’s a really great question. Because, yeah, these animals, they’re rodents, so they don’t sleep for eight hours at a time, right? So, how do you measure sleep deficiency? So, that was a hard question. We analyzed it several different ways. What we ended up with is using actigraphy. So, actigraphy is a non-invasive way of looking at rest and active states. So, these animals were video monitored. They were put into an actigraphy cage, which is an infrared beam that measures activity and rest. They were in that for their entire life. So, we took all of the wild type data.

We found that at the ages we were looking at, there were no changes. It was really, really robust with how many rest EPICs they had during their rest period, during the 12-hour rest period in their light-dark cycle. So, it was constant throughout every single age. So, we took that wild type value of what is standard rest. We compared it to the epileptic animals. We found that as they got closer to death that they had, they had more…

Sorry, I know my arm was not a great line graph. If this was wild type rest, every single day, there was very little variation. If this was the younger knockouts, they look just like the wild type and they had the same amount of rest. And then as they got closer to SUDEP, the rest started to go like this. So, the rest efficiency was this difference between the average all of the wild types together and then each knockout.

If you started treating chronically with DORA starting at, say, P30, and then did you happen to see it enhance survival in very large manner, say up to P20 or P50?

So that’s the study that we still need to do is give a DORA… There’s lots of different doors out there but give a DORA early on at a much younger age and then see if it can prevent everything from happening and if it can postpone SUDEP…

[I will] expand on that just a little bit is that for research purposes, if you give it before the onset of epilepsy, that’s not clinically relevant, right? So, you want to wait until some of the problems have started before you give a drug, so that it’s more clinically relevant. But I do think both of them need to be done, both studies need to be done, where you wait for the onset of some of the pathophysiologies to occur and then you give it as a treatment, but then you can actually start a treatment maybe earlier more as a proactive measure.

Now, it’s important to know that the DORAs improved sleep. So, you don’t want to give it during the period of the day that you’re supposed to be awake, because it’ll just knock you out and put you to sleep. So, this is really restricted to things that we can only do during the sleep. You can give it at lower doses to help protect against some of these pathologies, and it won’t induce sleep, but those are again nuances that we still need to figure out.

Are there any side effects to doing this chronic treatment with DORA?

So, DORAs are really safe. So, in 2009, when we first started working on these studies, we were using Almorexant, which was just about to be approved by the FDA. It has since been pulled from the market. Suvorexant and Lemborexant are now FDA use drugs. So, now we’re using drugs similar to those. So, the benzodiazepine and benzodiazepine-like drugs, when you give them as hypnotics, as a sleep aid drug, they do improve sleep, but they also can cause daytime drowsiness the next day. It can cause other kinds of side effects as well, like cognitive problems the next day and that kind of thing. So, the DORAs are really safe. They’re very well tolerated. You don’t end up with that next day drowsiness during the day.

So, in terms of their safety profile, clinically, these are very safe drugs so far. They’re still relatively new. So, let’s wait 5 to 10 years and see what happens. There is something interesting that we just recently found out. We have a study that’s actually under review right now, I didn’t have time to share this data. But we looked at sleep architecture and EEG of the epileptic animals that had been treated with many different anti-seizure drugs, traditional anti-seizure drugs.

So, in that study, we were able to separate out effects of a drug on sleep versus effects of a drug on a seizure. Because I know a lot of basic scientists and clinicians want to know, “Do the seizures arise more out of REM or non-REM?” We’re finding that for at least in our animal model, that the seizures arise both out of non-REM and out of REM just at baseline, and then the drugs change that up a little bit. Hopefully, that paper will be coming out soon.

What’s New at the ETSP?

Background:  The NINDS Epilepsy Therapy Screening Program (ETSP) is a preclinical screening program that provides researchers the opportunity to screen potential therapeutic agents in established rodent seizure models. Since its establishment, the ETSP has played a role in the development of several FDA-approved epilepsy drugs.

You Will Learn:  This webinar broadly discusses the scope of the ETSP and new models and assays available at the University of Utah.

Speaker Bio:  This webinar was presented by Dr. Cameron Metcalf, a Research Assistant Professor in Pharmacology and Toxicology at the University of Utah. He is also a Co-Investigator and the Associate Director of the Anticonvulsant Drug Development Program.  Dr. Metcalf’s primary research interests include the evaluation and advancement of novel therapies for the treatment of epilepsy and pain.

Download Full Transcript

Audience Q&A with Dr. Metcalf

In the intrahippocampal model, what areas do you record the EEG from?

So I would refer you to our colleagues at SynapCell for greater detail; they published their pharmacology and their methodology recently. But these recordings are deaf electrodes in the hippocampus. And since I don’t have any direct involvement with the actual performance of that model, I can’t say more than that. But I can refer you to their publications.

In the testing scheme which you showed at the beginning of your presentation, is it a step-wise process to go from the identification to the differentiation step? Or, how do you determine when a compound can move into the differentiation?

Sure. Thank you for the question. So one thing that I’ll say is that the decisions to advance compounds from one part of the testing scheme to another … we’re not always privy to those, and so I would defer to my colleagues at NINDS for specific examples.

What I can say generally is that the scheme is designed for it to be step-wise, meaning that as a compound advances through the identification phase, it would be a candidate to move toward the differentiation phase. What you might also envision is that while there are many compounds that could initially be tested in identification, not all those compounds have been advanced.

And so, there are some benchmarks and some go/ no-go decisions if you will, that help the ETSP in advancing compounds. There’s a lot of different factors that can come into play, and not all participants enter the program in equivalent ways. Some have to enter and exit for various reasons. Others may be limited in what their desires are to obtain information from the program. So I think it really varies, but in a very broad sense, it is intended to be step-wise.

Which amygdala nucleus is injected with kainic acid?

The model was designed around injection of the basolateral amygdala with kainic acid. And I should mention, this is also a good opportunity to highlight that we do have an upcoming manuscript that’s being prepared that will describe some of our methodologies.

Do you have any preliminary results in the intra-amygdala kainic acid model in mice, in terms of the seizure frequency and progression of the spontaneous seizures?

Yeah, great question, and again I would refer you to a paper that we will have forthcoming. We’ve also presented these at AES in the past. What I can say very broadly is that we do see variable seizure frequency similar to what I described for the rat kainate model. In a very general sense, we see on average about one seizure per day, but that really can vary quite dramatically. We can have animals that have a very high seizure burden, and we can have other animals that don’t have as many seizures, or seizures at all. And so we’re trying to find the best way to optimize our use of this model, and that’s going to be something that we’re going to continue to look at going forward.

When you do the tolerability toxicity assays, are you only working with models to look at motor activity, or are there other models that you look at for other types of toxicity?

Yeah, great question. This is something that we’ve thought a lot about recently. Historically, the models that we used to look at tolerability were really geared toward motor activity, and so that would include the rotarod assay in mice. We used a modified open field observational assay in rat. And we also used an open field automated locomotor open field assay.

However, in recent years, we’ve also brought on board a modified Irwin test. And while this isn’t as comprehensive as many you may be familiar with, it continues to be largely geared toward motor activity, but it also does allow us opportunity to look at other potential means of toxicity, such as autonomic, audio-visual and others. And so, this is helping us, but for those of you that work with rodent behavior, there’s only so much we can do, I think, to look at tolerability and be predictive of what happens in the clinical setting. But we have recently expanded our abilities to look at tolerability, particularly in rats.

What mice strain do you use for the acute seizure, the 6 Hz instead of epilepticus, because different strains can have drastic changes with regards to the severity of this?

Wonderful question, thank you, and this is also something we think a lot about. It does vary by model, so for 6 Hz and MES we use the CF-1 mouse strain from Charles River. For the corneal kindling assay, we have used the CF-1 animal in the past, however recently we’ve gone to the Charles River C57 black 6 model of mouse. For the TMEV assay, and for intra-amygdala kainate, we use C57s from Jackson. And so it does vary, but most directly to answer your question, we use the CF-1 mouse for our acute seizure assays.

But I do take the point that there are notable differences, not only in seizures but also in seizure pharmacology.