Beyond the Ion Channel | The ILAE Genetics Commission Blog
Epilepsies or seizure disorders are common diseases of the brain. Many types of epilepsies have a genetic etiology. Finding these genes and characterizing them will lead to insights about the physiology of epilepsies and -hopefully- novel treatment options. In our blog “Beyond the Ion Channel”, we try to make our research more understandable, digestable and interpretable.
Sodium channel. Voltage-gated channels for sodium ions are a crucial component of helping neurons depolarize and repolarize in a way that enables generation of action potentials. However, in order to function properly, voltage-gated ion channels co-exist in a fragile balance, and genetic alterations leading to functional changes in these channels are known causes of disease. SCN1A, SCN2A, and SCN8A have been implicated as causes for human epilepsy. However, SCN3A encoding the Nav1.3 channel, one of the most obvious candidates, could not be linked to disease so far. In a recent publication, we were able identify disease-causing mutations in this major neuronal ion channel. Interestingly, patients with an early onset and the most severe presentation had a prominent gain-of-function effect that responded to known antiepileptic medications.
Figure. Variants in SCN3A identified in our study (Zaman et al., 2018) and previous studies. Three variants found in four patients (p.I875T x2, p.P1333L, p.V1769A) were found to be de novo. A prominent gain-of-function was observed for p.Ile875Thr and p.Pro1333Leu. Both patients with the p.Ile875Thr had diffuse polymicrogyria.
Nav1.3. Voltage-gated sodium channels play a major role in tissues involved in the generation and propagation of electrical signals, most prominently the skeletal muscle, the heart muscle and, most importantly, the brain. There are nine different sodium channels in humans and four voltage-gated sodium channels expressed on neurons including the Nav1.1 (SCN1A), Nav1.2 (SCN2A), Nav1.3 (SCN3A), and Nav1.6 (SCN8A) channels. The naming of sodium channel is often a source for confusion between scientists involved in functional studies and genetic studies. Nav1.x refers to the protein, SCNxA to the gene, but for some sodium channel genes such as SCN8A, the numbering is off. Out of this quad of brain-specific sodium channels, three channels have been associated with epilepsy and neurodevelopmental disorders so far. SCN3A had been the missing sodium channel that was suggested in some previous studies, but never arrived at the same degree of certainty as the other sodium channels including SCN1A, SCN2A, and SCN8A. Nav1.3 encoded by the SCN3A gene is highly expressed in the developing brain, but its expression drops to basically insignificant levels postnatally.
Gain-of-function. In our recent study by Zaman and collaborators, we identified four patients with de novo mutations in SCN3A, including two patients with a recurrent variant. In contrast to previous studies, which indicated a role of SCN3A in milder neurodevelopmental disorders, the four patients described in our study had severe epilepsy starting in the first year of life. Functional studies identified a prominent gain-of-function effect of these mutations, indicating that increased channel function was the molecular mechanism behind the severe presentation in patients with SCN3A epileptic encephalopathy.
Nav1.3 expression. Looking at the expression pattern of SCN3A, this mutational mechanism is somewhat evident. For a gene with a very low expression after birth, only a significant increase in channel activity can result in hyperexcitability. Basically, in patients with SCN3A encephalopathy, the overactive Nav1.3 channel adds another sodium channel-related function that is not present in the typically developing brain – and the hyperexcitable network produces the epileptic encephalopathy. A related mechanism was shown for some variants in the SCN2A and SCN8A genes, as well, even though the mechanisms can only be compared to a certain degree. In vitro, the gain-of-function effect could be reduced by lacosamide and phenytoin, both commonly used anti-epileptic medications. This indicates that the identification of an underlying SCN3A mutation may have therapeutic consequences in some patients.
What you need to know. SCN3A de novo variants are a novel cause of epileptic encephalopathy. In patients with early-onset epileptic encephalopathies, the functional consequence of these mutations is a gain-of-function effect, resulting in hyperactive Nav1.3 channels that are most likely mediating the network hyperexcitability that results in epileptic encephalopathies. With the discovery of SCN3A as a new cause for epileptic encephalopathies, the Nav1.3 channel joins the other brain-specific sodium channels that are related to human epilepsy.
Ion channels and brain malformations. When the “channelopathy” concept first emerged – the idea that dysfunction of neuronal ion channels leads to neurological disease including epilepsy – it seemed implausible that such dysfunction could lead to malformations of cortical development. However, recent research has suggested that ion channel dysfunction may indeed be linked with brain malformations. In 2017, we saw convincing evidence that germline de novo variants in GRIN2B can cause malformations of cortical development. Some suggestive, but less conclusive, evidence has also linked SCN1A and SCN2A to brain malformations. Now Fry and collaborators demonstrate that de novo pathogenic variants in GRIN1 can also cause significant polymicrogyria, expanding the phenotypic spectrum of GRIN1-related disorders. As a disclaimer, I am also a co-author on the publication by Fry and collaborators.
Clustering of GRIN1 variants. A schematic diagram of the GRIN1 protein, showing the localization of polymicrogyria (PMG) associated GRIN1 variants on the top in red and non-PMG associated GRIN1 variants on the bottom in yellow. PMG-associated variants cluster in the M3 and S2 regions of the protein, which are important in channel gating and glycine binding. Over 50% of non-PMG associated variants are located in the transmembrane M4 helix, where no PMG-associated variants have been reported. Some GRIN1 variants have been reported in both patients with and without polymicrogyria. However, these patients have only had CT scans, which is not capable of detecting polymicrogyria.
GRIN1-related polymicrogyria. Fry and collaborators have identified de novo pathogenic variants in GRIN1 in 11 individuals with significant malformations of cortical development and associated neurodevelopmental disorders. MRI features included extensive bilateral cortical malformations, most consistent with polymicrogyria primarily in the frontal and parietal regions. One of the de novo GRIN1 variants was identified in a 22-week gestational age male fetus, with abnormal thinning and sulcation of the cerebral cortex, hypoplastic corpus callosum, and ventriculomegaly. This suggests that GRIN1-related cortical malformations may be detectable prenatally.
Gain-of-function vs. loss-of-function. Functional analysis polymicrogyria-associated GRIN1 variants identified an increased sensitivity to NMDA receptor agonists glycine and glutamate, consistent with a strong gain-of-function effect. This is in stark contrast to previously reported GRIN1 variants in patients without polymicrogyria, which were found to have significant dominant-negative and loss-of-function properties. Although this gain-of-function vs. loss-of-function dichotomy may be an oversimplification, it does suggest a correlation between the functional consequence of the GRIN1 variant and the resulting phenotype. Over-activation of the NMDA receptor may result in malformations of cortical development while loss of NMDA receptor function will result in a neurodevelopmental disorder without brain malformation. Although the mechanism whereby NMDA receptor gain-of-function results in neuronal migration defects is still not understood, one hypothesis is that the excitotoxic effects of NMDA receptor hyperactivation may lead to cell death during fetal brain development, which could lead to migrational defects in developing neurons.
Genotype-phenotype correlations. Comparison of GRIN1 variants in patients with polymicrogyria compared to patients without polymicrogyria showed that these variants cluster in different parts of the protein. The polymicrogyria-associated GRIN1 variants were highly clustered in the S2 domain and adjacent M3 helix regions of the GRIN1 protein. The S2 domain forms part of the glycine-binding domain. Glycine is an NMDA receptor activator. The polymicrogyria-associated variants in the M3 region were in a motif known to control NMDA receptor gating. This S2/M3 clustering of polymicrogyria-associated variants in GRIN1 is similar to what is seen in GRIN2B. In contrast, over 50% of previously reported GRIN1 variants in people without polymicrogyria are located in the M4 segment, where no polymicrogyria-associated variants were found.
Implications for precision medicine. It is likely too early to say whether or not these findings will directly influence precision medicine decisions for families affected by GRIN1-related disorders. However, the suggestion of genotype-phenotype correlations does provide additional evidence that may help in a clinical context. This research suggests that gain-of-function variants are most likely associated with a brain malformation phenotype whereas individuals with GRIN1-related disorders without brain malformations are more likely to have loss-of-function variants. Memantine, which is an NMDA receptor blocker, would theoretically be more useful for people with a confirmed gain-of-function variant. However, additional research is needed to investigate the usefulness of memantine in a clinical context, particularly for patients who have an underlying brain malformation.
Protocadherins. PCDH19-related epilepsy is the second most common genetic epilepsy, behind Dravet syndrome. PCDH19-related epilepsies display the unusual X-linked inheritance pattern in which heterozygous females are affected but hemizygous males are unaffected. Similarly, somatic mosaic males have also been reported. PCDH19 encodes protocadherin 19, a calcium-dependent cell-cell adhesion molecule that is highly expressed in the central nervous system. The long-hypothesized pathomechanism has been cellular interference, although experimental support has so far been lacking. Now, Pederick and collaborators provide evidence that supports the cellular interference mechanism in PCDH19-related epilepsies, bringing us closer to understanding the biology of this unusual genetic epilepsy.
PCDH19 mechanism. The differences in cell adhesion affinities lead to cellular interference in people with PCDH19-related epilepsies. In unaffected individuals (top panel), all cells in the developing brain have the same adhesion properties, leading to proper cell sorting and cell connections. In people with PCDH19-related epilepsy (middle panel), two cell populations exist: cells with wildtype PCDH19 and cells with variant PCDH19. These cells have different adhesion properties and therefore do not sort properly. Proper connections are not are not formed between the two cell populations. In males who carry a PCDH19 variant (bottom panel), a single homogeneous cell population exists. All of these cells have the same adhesion properties and therefore are sorted properly and have proper connections.
Cellular interference. The theory behind cellular interference is that two distinct cell populations exist in the developing brain of people with PCDH19-related epilepsies: cells containing unaltered PCDH19 and cells containing variant PCDH19. These two unique cell populations do not form proper connections with one another, leading to disrupted networks within the brain. However, in hemizygous males who only have one copy of PCDH19, a homogeneous cell population exists. Even though these cells carry a variant version of PCDH19, the fact that only a single cell population exists leads to the formation of proper connections. Therefore the existence of two types of cells – and not the PCDH19 variant itself per se – has been proposed to be the main driver of the phenotype in PCDH19-related epilepsies.
Experimental evidence for the mechanism. Protocadherins, including PCDH19, are adhesive molecules sticking out from the cell surface and can be conceptualized as the tiny hooks and loops of a strip of Velcro. PCDH19 is expressed in the brain with other types of protocadherins, and the various combinations of protocadherins determine a cell’s adhesive properties and thus how it is sorted during cortical development. Cells with similar adhesive properties are sorted together. Pathogenic variants in PCDH19 result in loss of adhesive function, which alters the types of cells the cell adheres to and the way it will be sorted during brain development.
Pederick and colleagues discovered that the brains of female mice who only had one functional copy of PCDH19 – equivalent to people with PCDH19-related epilepsy – showed distinct cell populations that were unable to adhere properly to one another. The cells expressing wildtype PCDH19 had normal adhesive properties, whereas the cells without PCDH19 had altered adhesive properties. The presence of two different adhesion affinities led to missorting of cells during brain development. This abnormality was not seen in male mice who completely lacked PCDH19. In the brains of male mice completely lacking PCDH19, neurons sorted appropriately because all cells exhibited the same adhesion affinities. The authors expand their study by further describing abnormal cortical sulcation patterns in four female patients with PCDH19-epilepsy, suggesting that subtle brain malformations may be a feature of this missorting process. However, due to random patterns of X-inactivation, the severity and presentations in people are likely highly variable.
What you need to know.Pederick and colleagues report evidence from mouse models of PCDH19-epilepsy that supports the cellular interference model. They demonstrate that the heterozygous cell populations – cells containing wildtype PCDH19 and cells containing altered PCDH19 – segregate abnormally during brain development. This alteration is not seen in male mice completely lacking PCDH19, which is consistent with what we see in humans. The authors further suggest that subtle brain malformations may be a part of the PCDH19 phenotype as an extreme consequence of this missorting process.
GAT1. When we first identified SLC6A1 in 2015, we were surprised that a significant proportion of patients with disease-causing variants in this gene had a rare epilepsy phenotype referred to as Myoclonic Astatic Epilepsy (MAE). Typically, at the time of gene discovery, it is often unclear how far the phenotypic spectrum expands. In a recent publication in Epilepsia, we reviewed the phenotype of 34 patients with SCL6A1-related epilepsy. Surprisingly, in contrast to many other epilepsy genes that showed a broad and occasionally non-specific phenotypic range, the SLC6A1-related phenotype expands beyond MAE, but remains centered around generalized epilepsies with a predominance of absence seizures and atonic seizures. It is a gene that has started to write its own story.
The SLC6A1 protein interaction network. The central role of SLC6A1 (GAT1) in GABA metabolism is best explained by reviewing the protein networks that it interacts with. On the one hand, SLC6A1 interacts with GAD1 and GAD2, the glutamate decarboxylases that turn glutamate into GABA. These proteins, in turn, interact closely with many presynaptic proteins including STXBP1, SNAP25, and VAMP2. On the other hand, GAT1 interacts with SLC1A2 and SLC1A3, transporters for excitatory amino acids such as glutamate.
SLC6A1. The SLC6A1 gene codes for the sodium- and chloride-dependent GABA transporter 1 and is the protein that removes GABA from the synaptic cleft. It is a central protein in GABA metabolism and is critical for regulating GABA, the main inhibitory neurotransmitter in the Central Nervous System. When we first described SLC6A1-related neurodevelopmental disorders, it became clear that many patients with disease-causing variants in this gene have epilepsy, specifically generalized epilepsies.
Paradoxical mechanism. Disease-causing variants in SLC6A1 are typically loss-of-function, suggesting that a reduction of GAT1 function is the disease mechanism. This is somewhat counterintuitive given the epilepsy phenotype – it remains unclear right now how the presumed excess of GABA in the synaptic cleft may result in hyperexcitability. But as with many other epilepsy genes, biological insights into these apparently contradictory mechanisms may hint at novel pathomechanisms.
The SLC6A1 phenotype. In our recent publication by Johannesen and collaborators, we review the phenotypes of 34 patients with disease-causing SLC6A1 variants including two families with several affected family members. Almost half of all patients had a phenotype compatible with Myoclonic Astatic Epilepsy (MAE), making SLC6A1 the gene with the closest association with an MAE phenotype to date. However, the phenotypic spectrum expanded significantly from our prior study. Seven other patients had other forms of generalized epilepsy, including Early Onset Absence Epilepsy (EOAE) and Eyelid Myoclonia with Absences (ELMA). We also identified two patients with focal epilepsies and three patients without seizures. With the exception of a single patient, all identified patients had some degree of intellectual disability. Mild to moderate intellectual disability with a focus on language development was the most commonly observed delay. In summary, the phenotypic spectrum of SLC6A1 is emerging as a relatively unique phenotypic range that is clearly distinct from other epileptic encephalopathies.
A unique spectrum. I would like to emphasize again how unusual it is to identify a gene with such an unusual phenotypic spectrum. Most genes for neurodevelopmental disorders have a broad phenotypic spectrum and usually span the various epilepsy subtypes including Infantile Spasms or other infantile epileptic encephalopathies. SLC6A1 does not seem to follow this pattern, but remains a gene for generalized epilepsies. This spectrum also points out a close relationship of the various rare generalized epilepsies that are typically distinct from the more common Idiopathic/Genetic Generalized Epilepsies (IGE/GGE) such as Childhood Absence Epilepsy (CAE) or Juvenile Myoclonic Epilepsy (JME). This suggest some common biological pathways in patients with MAE and conditions such as Early Onset Absence Epilepsy (EOAE) and Eyelid Myoclonia with Absences (ELMA) and is reminiscent of the phenotypic spectrum of the 15q13.3 microdeletion and 16p13.11 microdeletion.
What you need to know. In our recent study by Johannesen and collaborators, we describe the phenotypic spectrum in 34 patients with pathogenic variants in SLC6A1. The core phenotype of this gene is mild to moderate developmental delay with a generalized epilepsy that often presents as Myoclonic Astatic Epilepsy (MAE). Given this unusual phenotype, SLC6A1 emerges as one of the few genes for neurodevelopmental disorder with a relatively narrow phenotypic spectrum that sets the gene apart for the majority of other epilepsy genes.
Bomb Cyclone. While the entire US East Cost was held hostage by a weather system that introduced us to new catchy meterological concepts such as bombogenesis, I hope that everybody is staying warm and safe. I wanted to wish all our readers a Happy 2018 and try to give an outlook of the New Year in epilepsy genetics. Here are three things in epilepsy genetics that will happen in 2018 – and three things that won’t.
Bomb cyclone. The weather system covering the US East Coast with snow and a record-setting cold. The unusually strong phenomenon made news outlets go into overdrive by outdoing each other with novel superlatives referring to the lowest temperature recorded in Philadelphia in a generation or weather that is as cold as Mars (Image by NASA https://www.nasa.gov/multimedia/imagegallery/iotd.html).
1 – Appreciation of atypical phenotypes
When I look back at 2017 and ask myself what has impressed me the most clinically, I would emphasize the atypical phenotypes: milder presentations for seemingly severe genetic variants in neurodevelopmental genes, unusual neurological presentations for known disease entities, and unusual disease trajectories over time. As a clinician, I am fundamentally opposed to the idea that most genes for neurodevelopmental disorders results in a broad, undefined spectrum of neurological disorders and we are always on the lookout for novel phenotypic entities. My take-away from 2017 that we have only begun to understand the new phenotypic landscape that is opening up and is that our traditional boundaries may be shifting.
2 – A new phenotypic language
During our Special Interest Group at the American Epilepsy Congress (AES) in Washington, I used this slide to emphasize the emerging divide between the phenotypic language used in the clinical sphere and the language used in a diagnostic setting. While we typically think of epilepsy phenotypes with regards to the ILAE classification, this is not what most diagnostic laboratories use. Many diagnostic laboratories are involved in diagnostic testing of several diseases and increasingly used standardized vocabulary such as the Human Phenotype Ontology (HPO) for diagnostic purposes. We had initially developed the epilepsy-specific HPO terms within the first phase of our EuroEPINOMICS-RES project, but this concept has found little traction in the epilepsy field. Also, the ClinGen interface uses the MONDO disease ontology for disease classification that we’re currently aligning with our standard language in the epilepsy field. Take home message: we have to make sure that we don’t start talking in different diagnostic languages in 2018.
3 – Cloud platforms and their downstream effects
10,000 exomes? Just use your Google account to tap into the Cancer Genomic Cloud or possibly even some exomes on neurological disorders. Except that nobody is using these tools at the moment. While there was tremendous excitement for cloud-based platforms in the last two years, many of these tools have not gained the traction yet that was expected. Rather, we’re dealing with the downstream effects that we are realizing the start-up costs of using these tools that are often still limited to the major centers that host the bulk of the genetic data, anyway. What will happen in 2018? We will slowly accustom ourselves to large genetic data in the cloud and will develop a more mature view of what we can use this data for. The cloud is not the cure-all for all shortcoming in the genetics field, but another tool with advantages and disadvantages. Realizing the specific instances when such tools can be used, ranging from large-scale data mining to democratizing specific aspects of consortium-based research, will make them attractive for us in 2018.
Here are three things that won’t happen yet in 2018, but that are on the horizon for the future
1- Genetics meets MRI and EEG
Don’t get me wrong, I would love to see the “ultimate handshake”, the convergence of imaging/EEG research and genetics. However, this will still be a problem of numbers in 2018. The major issue: genetics requires larger cohorts for the association approach and to overcome genomic noise. Imaging/EEG research, in contract, is deep phenotyping on individuals. Once we have a cohort of 1000+ patient with existing genetic and imaging/EEG data, we can start looking. Everything else is probably still underpowered and not achievable in 2018.
2 – Making sense of electronic medical records
As many of you already know, using information from electronic medical records in patients with known or presumed genetic epilepsies has become a major focus of our group. The ultimate goal is to overcome the phenotypic bottleneck that is still presenting one of the major hold-ups in epilepsy genetics. I am actually very excited by the idea that we can use readily available data in the medical records for computational and longitudinal phenotyping for concepts such as drug response, side effects, and developmental outcomes. In fact, within our GRIN collaboration, we might currently have one of the largest pediatric datasets, encompassing more than 1,000 “exome years” (available exomes x patient years in the electronic medical records). However, the more we handle the data, the more we realize that we have to be careful and rigorous when assessing some of the basic assumptions. Take genetic testing as an example – if we ask for the most significant findings in a genome-wide association studies without any correction, we’ll mainly find one thing: artefacts. Electronic medical records for computational phenotypic hold great promise, but this novel field requires some basic work. The big promise is the prospect of “scalability” – once the basic rules have been established, phenotyping can grow as exponentially as genetic testing – but 2018 will probably be reserved for doing some of the basic footwork.
3 – Meaningfully implementing patient-derived data
If Facebook can do it, why can’t we? Conceptually, it should be possible to use the already existing technology to use patient-derived data for improving our understanding of genetic epilepsies. This may be through patient-owned seizure diaries or through wearables. However, in 2017, I have grown slightly skeptical that a breakthrough on this area is on the horizon. I started out 2017 being quite excited that patient-owned seizure diaries will lead to an explosion of available data that will help us better understand epilepsy. Technically, the tools are available, but there is a remaining disconnect that we were not able to fully address. None of the tools that I am aware of has found a major traction in the patient community. Please don’t understand me wrong – I’m not blaming the community, I’m blaming the tools. The existing tools and providers have not been able to sufficiently convince the patient community that data sharing is both safe and meaningful, which leaves a tremendous opportunity to improve our understanding of genetic epilepsies untapped. I still believe that if we are able to capture some of the dynamics on social media such as Facebook for scientific purposes, we may learn about aspects of genetic epilepsies that we’re currently oblivious to. However, I don’t think that this will be an achievable goal for 2018 yet.
Controversies. While you are packing your bags for the 71st Annual Meeting of the American Epilepsy Society in Washington, D.C., we wanted to point out one agenda item that may be of interest for you. The AES agenda typically has many parallel sessions, so I wanted to make a plug for our Genetics Special Interest Group (SIG) on Friday, 12/1 at 1:30PM. The topic of our SIG is going to be “Making Sense of Genetic Data in Epilepsy – Consensus and Controversy in 2017”. In contrast to regular sessions and lectures, a SIG is meant to stimulate discussion between SIG members. Therefore, in parallel to previous years, we would like to invite the attendees to use the opportunity to discuss challenging cases within a dedicated AES Special Interest Group.
Confocal microscopy of mouse cortex (modified under a CC licence from https://www.flickr.com/photos/zeissmicro/10799674936)
The 2017 SIG. During the current SIG, we will have three speakers giving us a brief introduction to particularly challenging aspects of epilepsy genetics/precision medicine. Heather Mefford will discuss issues about novel genes and gene validity. Afterwards, we have invited Katie Helbig to talk about challenges in variant interpretation including the current issues surrounding the limitations of current guidelines and genetic literacy. Finally, Mark Fitzgerald will introduce challenging issues surrounding the implementation of genetic findings into clinical practice, including some of the more recent issues surrounding precision treatment with compounds such as quinidine. Again, this is not a lecture, but a Special Interest Group. It is THE place at AES to focus on the controversies surrounding these issues. There is a reason why we entitled the SIGs “Variant Fight Club” at past AES meetings.
Bring your own variant. If you would like to run challenging cases or scenarios relating to epilepsy genetics and precision medicine by us during the SIG, feel free to contact us in advance. We don’t need much information, a brief email to firstname.lastname@example.org will do. Looking forward to seeing you on Friday! The room for the SIG is 102 B, Street Level in the Convention Center.
KCNA2. We have previously discussed KCNA2 and that pathogenic variants in this gene can lead to a spectrum of neurological phenotypes. Pathogenic KCNA2 variants were first recognized in individuals with early-onset developmental and epileptic encephalopathies and have subsequently been found also in individuals and families with hereditary spastic paraplegia, episodic ataxia, and milder epilepsies. KCNA2 encodes the Kv1.2 potassium channel, a delayed rectifier class of potassium channel that enables neuronal repolarization after an action potential. A new study by Masnada and colleagues provides clinical and functional data from 23 patients, representing the largest KCNA2 cohort reported to date. Within the KCNA2-related encephalopathy spectrum, it now seems that there may be three distinct phenotypes.
Figure. Schematic figure of Kv1.2 potassium channel with distribution of de novo KCNA2 variants. Red circles represent loss-of-function variants, yellow circles represent gain-of-function variants, and blue circles represent gain- and loss-of-function variants. Prominent recurrent de novo variants from each functional class are designated: R297Q (gain-of-function), T374A (gain- and loss-of-function), and P405L (loss-of-function).
Three functional classes. When we first wrote about KCNA2, we discussed the activating and abolishing categories of variants. The recent work by Masnada and colleagues expands and complicates this functional picture. The abolishing (loss-of-function) and activating (gain-of-function) classes of variants remain, but a third functional class emerges that shows a combination of activating and abolishing properties (gain- and loss-of-function). The loss-of-function variants lead to a loss of channel activity with a dominant-negative effect on wildtype subunits. The gain-of-function variants lead to permanently open channels. However in the third functional class, the observed gain-of-function effect is reduced by an additional loss-of-function, through a variety of mechanisms including shifts in steady state activation and decreased current amplitude.
The three phenotypes. Masnada and colleagues propose that KCNA2-encephalopathy is a spectrum that consists of three clinical phenotypes, which correspond to the effect of the underlying KCNA2 variant on Kv1.2 channel function: (1) loss-of-function, (2) gain-of-function, and (3) gain- and loss-of-function. All three phenotypic groups have overlapping features, including seizure onset within the first year of life, fever-sensitive seizures, developmental and cognitive impairment, behavioral issues, and signs of cerebellar involvement such as ataxia, coordination difficulties, and dysarthria. However each of the three groups has distinctive features. The loss-of-function group showed predominant focal seizures compared to predominant generalized seizure in the gain-of-function group. EEG findings in the loss-of-function group showed activation of epileptiform activity during non-REM sleep, which raised the concern for ESES in some patients. Although ataxia was observed in all three groups, it was most severe in the two gain-of-function groups. MRI was normal in the loss-of-function group, but both gain-of-function groups showed progressive cerebellar atrophy, beginning in childhood in the gain- and loss-of-function group. Developmental outcomes appear to be more severe in the two gain-of-function groups compared to the loss-of-function group. The most severe phenotypes were observed among patients with gain- and loss-of-function variants, who often had neonatal-onset epilepsy with severe to profound intellectual disability.
Variant specific effects. It is worth noting that three recurrent variants accounted for two-thirds of the patients in the study. The P405L loss-of-function variant was found in 5 individuals and accounts for 5/8 (63%) patients in the loss-of-function group. The R297Q gain-of-function variant was found in 7 individuals and accounts for 7/9 (78%) patients in the gain-of-function group. The T374A gain- and loss-of-function variant was found in 3 individuals and accounts for 3/6 (50%) patients in the gain- and loss-of-function group. Taking into account the high proportion of each of these three variants in the three respective functional/phenotypic classes, it is reasonable to wonder whether we are observing variant-specific effects rather than functional class-specific effects. For example, patients with the P405L variant present with a relatively homogeneous clinical picture, including febrile, focal dyscognitive, and hemiclonic seizures, with remission in late childhood or early adolescence, and an EEG picture reminiscent of ESES. Is the emerging clinical picture of the loss-of-function class of variants due to loss of Kv1.2 function in general, or is there something specific about the P405L variant itself that is leading to this clinical picture? Additionally, the severe phenotype associated with gain- and loss-of-function variants may be highly influenced by the T374A variant, which seems to be associated with a particularly severe outcome including neonatal-onset seizures, spastic quadriplegia, cerebellar atrophy, and profound ID. The remaining three patients in the gain- and loss-of-function group do not appear to be as severely affected as the three who carry the T374A variant. Further studies of larger patient cohorts are required before we can confidently assess whether KCNA2-related phenotypes are determined based on functional class or are variant-specific.
Synaptic transmission. Over the last several years, pathogenic variants in multiple genes involved in synaptic transmission have been identified in early-onset epilepsies. STXBP1 and STX1B both encode components of the SNARE complex, a complicated protein complex that mediates the fusion of the plasma membrane of the presynapse and the synaptic vesicle to allow for neurotransmitter release. DNM1, encoding the dynamin-1 protein, plays an essential role in recycling synaptic vesicles back into the presynapse after neurotransmitter release. A new study by Myers and collaborators has identified several patients with de novo variants in PPP3CA, which encodes another protein involved in synaptic vesicle recycling, further highlighting the importance of synaptic transmission in the etiology of severe neurodevelopmental disorders. In the interest of full disclosure, I am also a co-author on this study.
Figure. PPP3CA encodes the catalytic A subunit of the enzyme calcineurin shown above in yellow. Calcineurin is a calcium dependent serine-threonine phosphatase that modulates dynamin-1 mediated synaptic vesicle recycling. After nerve depolarization, calcineurin responds to an increase in calcium ions in the presynapse and removes the phosphate group (red) from inactive dynamin-1 (purple). The now dephosphorylated dynamin-1 is active and is able to participate in synaptic vesicle fission.
Calcineurin and dynamin-1 interaction. PPP3CA encodes the catalytic A subunit of calcineurin, a calcium- and calmodulin-dependent serine/threonine phosphatase. Calcineurin is known to play an important role in T-cell activation. Many readers may recognize calcineurin because it is the target of calcineurin inhibiting drugs, including cyclosporin. However, calcineurin plays another critical role in the brain and is responsible for the calcium-dependent dephosphorylation of dynamin-1 after nerve depolarization. As we have discussed previously on this blog, DNM1 is a critical protein in vesicle fission. Dynamin-1 proteins form a functional clamp around the base of the synaptic vesicle, fused to the plasma membrane, after neurotransmitter release. This dynamin clamp then twists tighter and tighter until the vesicle pops off and is recycled back into the presynapse. Calcineurin plays a crucial role in this process; in a phosphorylated state, dynamin-1 remains inactive in the cytoplasm. However, after nerve depolarization, calcineurin dephosphorylates dynamin-1, kicking it into its active state.
The PPP3CA phenotype. What is the clinical picture of these patients, and does it resemble DNM1 encephalopathy? Myers and collaborators identify six unrelated patients with de novo PPP3CA variants. Five of these individuals came from a cohort of over 4,700 individuals with neurodevelopmental disorders, which is significantly more de novo variation in PPP3CA than would be expected by chance. All individuals have severe neurodevelopmental disorders; five out of six have epilepsy. Although the clinical picture is in some ways similar to DNM1 encephalopathy – all affected individuals have severe to profound intellectual disability and 3/6 presented with a developmental and epileptic encephalopathy picture – the phenotype of PPP3CA is less homogeneous than what we have seen so far in DNM1, where most patients described so far have infantile spasms and/or Lennox-Gastaut syndrome. The epilepsy phenotype associated with pathogenic variants in PPP3CA, at least based on these five patients, was more variable with no consistent pattern emerging in terms of age of onset, frequency, or type of seizures observed.
Functional insights. The functional mechanism of the identified PPP3CA variants remains to be elucidated. However, it is notable that a recurrent variant (Gln282Lys) was reported in two individuals. The active site of PPP3CA is coordinated by three histidine amino acid residues (His92, His199, His281), one asparagine (Asn150), and two aspartic acids (Asp90, Asp118). Four of the six reported individuals in this study have de novo variants affecting these or adjacent amino acid residues, suggesting that these variants may interfere with the enzymatic properties of the protein.
What you need to know. Myers and collaborators have identified six unrelated individuals with de novo missense variants in PPP3CA, which encodes the catalytic A subunit of calcineurin. Calcineurin is responsible for calcium-dependent dephosphorylation of dynamin-1 and therefore plays an important role in synaptic vesicle recycling. This study implicates another synaptic transmission-related gene in the etiology of severe neurodevelopmental disorders, further emphasizing the biological importance of this pathway in neurodevelopment.
Epilepsiome, meet ClinGen. For more than a year, I have meant to write about the extension of the Epilepsiome effort to our ClinGen epilepsy working group. The overall ClinGen framework is a NIH-funded resource dedicated to building a central resource that defines the clinical relevance of genes and variants for use in precision medicine and research. Within this framework, the ClinGen Epilepsy Working group is a group of curators to apply the formal framework to epilepsy genes. Given the explosion of genetic data, curating epilepsy genes is important as a basis for precision medicine and long overdue. Within our epilepsy working group, we build upon the ClinGen framework to make it applicable to epilepsy genes. Here is what you need to know about epilepsy gene curation.
Columbus, Ohio. During my visit to Columbus, Ohio as an invited speaker at the National Society of Genetic Counselors (NSGC), we had dinner in the Leveque tower. The tower once served as an aerial lighthouse to Amelia Earheart during her pioneer flights. For some reason, the light tower analogy stuck with me when thinking about our process to navigate through the jungle of epilepsy genes, trying to figure out the evidence for genes and how to conceptualize phenotoypes and functional evidence within our ClinGen framework (from https://www.flickr.com/photos/sentfromthepast/36194305014 under a CC BY-NC 2.0 licence)
Columbus, Ohio. During the annual meeting of the National Society of Genetic Counselors in Columbus, Ohio, I was able to briefly meet up with Erin Riggs. Erin coordinates our epilepsy working group that is co-chaired by Heather Mefford and myself. We used this opportunity to discuss our working group and our way forward and this discussion was motivation for me to sit down and put this blog post into words that I have been thinking about for so long.
Gene curation. Gene curation basically means that we try to assess the evidence that a given gene is related to epilepsy. In ClinGen, the formal process includes a literature review including the review of genetic evidence and functional evidence, followed by an assessment of this evidence in a predefined matrix. Genes can accrue points for positive evidence, which will then add up to classifying the gene with limited, moderate, or strong evidence. If a gene holds up over time, it becomes a definite gene. If there is contradictory evidence, a gene can become disputed or eventually even refuted. A work-up for any given gene is time consuming and we are currently supported by a motivated and talented gene curator team as we work through the first genes.
Epilepsy gene curation. The role of working groups is to be trained in the formal ClinGen assessment and adapt the formal ClinGen framework to the disease-specific area. It is our role to understand what specific rules we need for epilepsy. For example, during our initial curation work, we already realized that that we need clear guidelines on what is called “precuration” – what is the phenotype that we are actually curating for. For example, should the SCN1A gene be curated for separate phenotypes or for a spectrum of phenotypes. It has already become clear that reconciling our clinical thinking with the overall ClinGen phenotyping framework is not without difficulties.
Phenotypic overlap. The “phenotypic” overlap in the gene curation process is one of the major challenges in diagnostic reports and there are currently no standards for this. This may lead to over- or under-interpretations of variants of uncertain significance depending on how the phenotypic overlap is perceived. The other challenge that we have encountered is the evaluation of functional evidence and animal models. How do we score a knock-out mouse that has ataxia, but not seizures? Should this count towards the evidence for a gene or not count at all? While we are finishing our first genes, we slowly zero in on realistic rules to handle the most common questions in the curation process.
Why curation is important. There are large databases to tell us whether variants in genes have been identified previously and great variant predictions tools that have vastly improved our ability to predict whether a certain variant is benign or possible contributing. While the final decision whether a variant within a given gene is causative for a given disease ultimately cannot be automated, this decision-making process can be supported by the robust frameworks that are being developed. For example, the review of a gene’s PLI -indicating a gene’s intolerance towards loss-of-function variants- is a standard in our group. Gene curation, the critical review of the evidence for a given gene, is finally at a point where we have standards and ways that variant prediction can facilitate some of our work. Epilepsy genetic testing is an expanding field and not all genes that are tested for on a diagnostic basis are valid epilepsy genes. Hence, a clear dictionary of genes and their relationship to epilepsy is required to facilitate the diagnostic process and reduce the number of false positive and false negative genetic findings in patients with epilepsy.
Epilepsiome to ClinGen. I am currently completing my 3-year term on the ILAE Genetics Commission and we have slowly moved over the Epilepsiome effort to our ClinGen group, which now provides the new home to our gene curators. Many of our curators started with us in the Epilepsiome and the ClinGen epilepsy group is happy to accept further members to help us tackle the growing body of epilepsy genes. I would like to use this opportunity to thank our “guardians of the epilepsy genes” for their ongoing effort and to acknowledge the ILAE for their support in launching the Epilepsiome initiative.
Neuronal spectrinopathies. Spectrins are a major component of the neuronal cytoskeleton, the scaffold underneath the cell membrane that gives cells their characteristic shape and anchors transmembrane proteins such as voltage-gated ion channels. SPTAN1, the gene coding for the non-erythrocyte alpha-II spectrin, has been known as a rare cause of early-onset epileptic encephalopathies with hypomyelination and atrophy. However, the full phenotypic spectrum and the range of pathogenic variants was unknown. In a recent publication in Brain, 20 patients with pathogenic variants in SPTAN1 are reported, expanding the known range of phenotypes and suggesting a very unusual disease mechanism through in-frame deletions or duplications. Here is what links the neuronal cytoskeleton to epileptic encephalopathies.
Figure 1. Neuronal spectrins and the cytoskeleton. (A) the cytoskeleton of the axon is supported by spectrin tetramers intersperse with actin rings. This regular structure can be observed in cell cultures and forms in the first week in vitro (DIV days in vitro). Spectrins are also important in providing the shape of red blood cells, even though the specific subtypes are different (from Stuart and Chen eLIFE 2015, DOI: 10.7554/eLife.06235 under a CC license). (B) Spectrins form tetramers between alpha and beta spectrins. Ankyrin (ANK) provides the connection with ion channels in the axonal membrane. The current cartoon demonstrates the position of disease-related pathogenic variants in some patients with SPTAN1 coding for an alpha-spectrin and SPTBN2, coding for a beta spectrin. Most pathogenic variants are found in the dimerization site where SPTAN1 variants have been shown to be dominant negative (from Lise et al., 2012 https://doi.org/10.1371/journal.pgen.1003074 under a CC license)
Spectrin. In order to acquire a certain shape, cells require a sophisticated scaffold underneath the cellular membrane. This is particularly true for cells like neurons, where the cellular shape branching into potentially hundreds of dendritic and axonal branches, is tightly linked to the specific neuronal function. For example, spinocerebellar ataxia 5 (SCA5) or “Lincoln ataxia” is due to pathogenic variants in the SPTBN2 gene, coding for a neuronal beta-III spectrin. Spinocerebellar ataxias are neurodegenerative disorders arising from the degeneration from neuronal pathways. Given the critical importance of the cytoskeleton, it makes intuitive sense that variants in spectrin genes may results in the degeneration of neurons over time. However, the discovery of SPTAN1 in patients with early-onset epileptic encephalopathy, hypomyelination, and progressive atrophy of the cortex, brainstem, and cerebellum questioned our understanding of how the cytoskeleton may be involved in neurological disorders.
Mutational spectrum. In the publication by Syrbe and collaborators, 20 patients with pathogenic SPTAN1 variants are described. The disease spectrum comprises the classical SPTAN1 phenotype, but also includes patients with mild to moderate intellectual disability. Interestingly, patients with variants in the end of the SPTAN1 gene, which codes for the heterodimer region, appear to have the most profound disease presentation. This particular region within the alpha-2 spectrin is important for connecting the alpha-spectrins with beta-spectrins and pathogenic variants within this region have previously been shown to be dominant-negative – in summary, this domain is so critical that any disruption to the protein helices not only makes the single copy of alpha-spectrin non-functional, it also disrupts the generation of spectrin tetramers to a degree that is more worse than a truncation in the SPTAN1 gene.
In-frame. The SPTAN1 gene is one of the few genes where in-frame deletions and duplications have been shown to be pathogenic. In the study by Syrbe and collaborators, nine in-frame deletions were identified compared to seven missense variants. This means that more than half of the disease-causing variants in SPTAN1 are due to in-frame deletions and duplications. Typically, in-frame deletions or duplications in other genes would be considered variants of uncertain significance and are often difficult to interpret; these types of variants may even be filtered out as benign variants by some analysis pipelines. In-frame variants are either deletions or duplications by three or a multitude of three base pairs, leaving reading frame of the DNA intact, but adding or deleting single or multiple amino acids. The non-erythrocytic alpha-2 spectrin SPTAN1 is unusual as it does not tolerate these in-frame changes in the region that links different spectrins together. This region requires such a well-defined amino acid sequence that even the interjection of a single amino acid may break the connection between proteins. The authors investigated some SPTAN1 variants in patient fibroblasts and demonstrate that these variants lead to spectrin aggregates – the altered spectrin is stuck in the cell and cannot properly function as a scaffolding protein.
What you need to know. The phenotypic spectrum of SPTAN1 encephalopathy is wider than previously known and particularly patients with variants outside the heterodimer region can have milder phenotypes. More than half of the variants affecting the heterodimer region are in-frame variants, an unusual disease mechanism that likely has to do with interfering with the complex structure that spectrins provide. Some of these in-frame variants are recurrent and the particular p.(Aps203_Leu2305dup) three-amino acid duplication was found in 5 patients.
Read Full Article
Read for later
Articles marked as Favorite are saved for later viewing.
Scroll to Top
Separate tags by commas
To access this feature, please upgrade your account.