HCN1 epilepsy: From genetics and mechanisms to precision therapies

Pathogenic variation in HCN1 is now an established cause of epilepsy and intellectual disability. Variation in HCN1 causes a spectrum of disease with a genotype-phenotype relationship emerging. De novo pathogenic variants that occur in the transmembrane domains of the channel typically cause a cation 'leak' that associates with severe developmental and epileptic encephalopathy (DEE). Genotype-phenotype associations for variants that fall outside of the transmembrane domains are less well established but do include milder forms of epilepsy that can be either de novo or inherited. HCN1 DEE mouse models have been generated which recapitulate the seizures and learning difficulties seen in human patients. These mice have also acted as powerful preclinical models which share pharmacoresponsiveness with human HCN1 DEE patients. Data from these mouse models support the conclusion that anti-seizure medications with sodium channel block as their primary mechanism of action should be used with caution in HCN1 DEE. Other comorbidities of HCN1 DEE including retinal dysfunction have also been modelled in HCN1 DEE mice, suggesting HCN1 variants can cause a dramatically reduced sensitivity to light with limited ability to process temporal information. Our understanding of the genetics and pathophysiological mechanisms underlying HCN1 epilepsy has progressed significantly and is already influencing therapy. However, more research effort is needed to fully understand the natural histories of HCN1 epilepsies and to develop precision therapeutic approaches.

Cation leak: a common functional defect causing HCN1 developmental and epileptic encephalopathy

Pathogenic variants in HCN1 are an established cause of developmental and epileptic encephalopathy (DEE). To date, the stratification of patients with HCN1-DEE based on the biophysical consequence on channel function of a given variant has not been possible. Here, we analysed data from eleven patients carrying seven different de novo HCN1 pathogenic variants located in the transmembrane domains of the protein. All patients were diagnosed with severe disease including epilepsy and intellectual disability. The functional properties of the seven HCN1 pathogenic variants were assessed using two-electrode voltage-clamp recordings in Xenopus oocytes. All seven variants showed a significantly larger instantaneous current consistent with cation leak. The impact of each variant on other biophysical properties was variable, including changes in the half activation voltage and activation and deactivation kinetics. These data suggest that cation leak is an important pathogenic mechanism in HCN1-DEE. Furthermore, published mouse model and clinical case reports suggest that seizures are exacerbated by sodium channel blockers in patients with HCN1 variants that cause cation leak. Stratification of patients based on their 'cation leak' biophysical phenotype may therefore provide key information to guide clinical management of individuals with HCN1-DEE.

Retinal Dysfunction in a Mouse Model of HCN1 Genetic Epilepsy

Pathogenic variants in HCN1 are associated with a range of epilepsy syndromes including a developmental and epileptic encephalopathy. The recurrent de novo HCN1 pathogenic variant (M305L) results in a cation leak, allowing the flux of excitatory ions at potentials where the wild-type channels are closed. The Hcn1M294L mouse recapitulates patient seizure and behavioral phenotypes. As HCN1 channels are highly expressed in rod and cone photoreceptor inner segments, where they shape the light response, mutated channels are likely to impact visual function. Electroretinogram (ERG) recordings from male and female mice Hcn1M294L mice revealed a significant decrease in the photoreceptor sensitivity to light, as well as attenuated bipolar cell (P2) and retinal ganglion cell responses. Hcn1M294L mice also showed attenuated ERG responses to flickering lights. ERG abnormalities are consistent with the response recorded from a single female human subject. There was no impact of the variant on the structure or expression of the Hcn1 protein in the retina. In silico modeling of photoreceptors revealed that the mutated HCN1 channel dramatically reduced light-induced hyperpolarization, resulting in more Ca2+ flux during the response when compared with the wild-type situation. We propose that the light-induced change in glutamate release from photoreceptors during a stimulus will be diminished, significantly blunting the dynamic range of this response. Our data highlight the importance of HCN1 channels to retinal function and suggest that patients with HCN1 pathogenic variants are likely to have a dramatically reduced sensitivity to light and a limited ability to process temporal information.SIGNIFICANCE STATEMENT Pathogenic variants in HCN1 are emerging as an important cause of catastrophic epilepsy. HCN1 channels are ubiquitously expressed throughout the body, including the retina. Electroretinogram recordings from a mouse model of HCN1 genetic epilepsy showed a marked decrease in the photoreceptor sensitivity to light and a reduced ability to respond to high rates of light flicker. No morphologic deficits were noted. Simulation data suggest that the mutated HCN1 channel blunts light-induced hyperpolarization and consequently limits the dynamic range of this response. Our results provide insights into the role HCN1 channels play in retinal function as well as highlighting the need to consider retinal dysfunction in disease caused by HCN1 variants. The characteristic changes in the electroretinogram open the possibility of using this tool as a biomarker for this HCN1 epilepsy variant and to facilitate development of treatments.

Efficacy of antiseizure medication in a mouse model of HCN1 developmental and epileptic encephalopathy

Acquisition of drug-sensitivity profiles is challenging in rare epilepsies. Anecdotal evidence suggests that antiseizure medications that block sodium channels as their primary mechanism of action exacerbate seizures in HCN1 developmental and epileptic encephalopathies (DEEs), whereas sodium valproate is effective for some patients. The Hcn1 M294L heterozygous knock-in (Hcn1M294L ) mouse carries the homologue of the recurrent gain-of-function HCN1 M305L pathogenic variant and recapitulates the seizure and some behavioral phenotypes observed in patients. We used this mouse model to study drug efficacy in HCN1 DEE. Hcn1M294L mice display epileptiform spiking on electrocorticography (ECoG), which we used as a quantifiable measure of drug effect. Phenytoin, lamotrigine, and retigabine significantly increased ECoG spike frequency, with lamotrigine and retigabine triggering seizures in a subset of the mice tested. In addition, there was a strong trend for carbamazepine to increase spiking. In contrast, levetiracetam, diazepam, sodium valproate, and ethosuximide all significantly reduced ECoG spike frequency. Drugs that reduced spiking did not cause any consistent ECoG spectral changes, whereas drugs that increased spiking all increased power in the slower delta and/or theta bands. These data provide a framework on which to build our understanding of gain-of-function HCN1 DEE pharmacosensitivity in the clinical setting.

Cation leak underlies neuronal excitability in an HCN1 developmental and epileptic encephalopathy

Pathogenic variants in HCN1 are associated with developmental and epileptic encephalopathies. The recurrent de novo HCN1 M305L pathogenic variant is associated with severe developmental impairment and drug-resistant epilepsy. We engineered the homologue Hcn1 M294L heterozygous knock-in (Hcn1M294L) mouse to explore the disease mechanism underlying an HCN1 developmental and epileptic encephalopathy. The Hcn1M294L mouse recapitulated the phenotypic features of patients with the HCN1 M305L variant, including spontaneous seizures and a learning deficit. Active epileptiform spiking on the electrocorticogram and morphological markers typical of rodent seizure models were observed in the Hcn1M294L mouse. Lamotrigine exacerbated seizures and increased spiking, whereas sodium valproate reduced spiking, mirroring drug responses reported in a patient with this variant. Functional analysis in Xenopus laevis oocytes and layer V somatosensory cortical pyramidal neurons in ex vivo tissue revealed a loss of voltage dependence for the disease variant resulting in a constitutively open channel that allowed for cation 'leak' at depolarized membrane potentials. Consequently, Hcn1M294L layer V somatosensory cortical pyramidal neurons were significantly depolarized at rest. These neurons adapted through a depolarizing shift in action potential threshold. Despite this compensation, layer V somatosensory cortical pyramidal neurons fired action potentials more readily from rest. A similar depolarized resting potential and left-shift in rheobase was observed for CA1 hippocampal pyramidal neurons. The Hcn1M294L mouse provides insight into the pathological mechanisms underlying hyperexcitability in HCN1 developmental and epileptic encephalopathy, as well as being a preclinical model with strong construct and face validity, on which potential treatments can be tested.

Loss-of-function variants in Kv 11.1 cardiac channels as a biomarker for SUDEP

Objective

To compare the frequency and impact on the channel function of KCNH2 variants in SUDEP patients with epilepsy controls comprising patients older than 50 years, a group with low SUDEP risk, and establish loss‐of‐function KCNH2 variants as predictive biomarkers of SUDEP risk.

Methods

We searched for KCNH2 variants with a minor allele frequency of <5%. Functional analysis in Xenopus laevis oocytes was performed for all KCNH2 variants identified.

Results

KCNH2 variants were found in 11.1% (10/90) of SUDEP individuals compared to 6.0% (20/332) of epilepsy controls (p = 0.11). Loss‐of‐function KCNH2 variants, defined as causing >20% reduction in maximal amplitude, were observed in 8.9% (8/90) SUDEP patients compared to 3.3% (11/332) epilepsy controls suggesting about threefold enrichment (nominal p = 0.04). KCNH2 variants that did not change channel function occurred at a similar frequency in SUDEP (2.2%; 2/90) and epilepsy control (2.7%; 9/332) cohorts (p > 0.99). Rare KCNH2 variants (<1% allele frequency) associated with greater loss of function and an ~11‐fold enrichment in the SUDEP cohort (nominal p = 0.03). In silico tools were unable to predict the impact of a variant on function highlighting the need for electrophysiological analysis.

Interpretation

These data show that loss‐of‐function KCNH2 variants are enriched in SUDEP patients when compared to an epilepsy population older than 50 years, suggesting that cardiac mechanisms contribute to SUDEP risk. We propose that genetic screening in combination with functional analysis can identify loss‐of‐function KCNH2 variants that could act as biomarkers of an individual’s SUDEP risk.

Are Variants Causing Cardiac Arrhythmia Risk Factors in Sudden Unexpected Death in Epilepsy?

Sudden unexpected death in epilepsy (SUDEP) is the most common cause of premature mortality in individuals with epilepsy. Acute and adaptive changes in heart rhythm in epilepsy implicate cardiac dysfunction as a potential pathogenic mechanism in SUDEP. Furthermore, variants in genes associated with Long QT syndrome (LQTS) have been identified in patients with SUDEP. LQTS is a cardiac arrhythmia condition that causes sudden cardiac death with strong similarities to SUDEP. Here, we discuss the possibility of an additive risk of death due to the functional consequences of a pathogenic variant in an LQTS gene interacting with seizure-mediated changes in cardiac function. Extending this general concept, we propose a hypothesis that common variants in LQTS genes, which cause a subtle impact on channel function and would not normally be considered risk factors for cardiac disease, may increase the risk of sudden death when combined with epilepsy. A greater understanding of the interaction between epilepsy, cardiac arrhythmia, and SUDEP will inform our understanding of SUDEP risk and subsequent potential prophylactic treatment.

The hyperpolarization-activated cyclic nucleotide-gated 4 channel as a potential anti-seizure drug target

BACKGROUND AND PURPOSE

Hyperpolarization-activated cyclic nucleotide-gated (HCN) channels are encoded by four genes (HCN1-4) with distinct biophysical properties and functions within the brain. HCN4 channels activate slowly at robust hyperpolarizing potentials, making them more likely to be engaged during hyperexcitable neuronal network activity seen during seizures. HCN4 channels are also highly expressed in thalamic nuclei, a brain region implicated in seizure generalization. Here, we assessed the utility of targeting the HCN4 channel as an anti-seizure strategy using pharmacological and genetic approaches.

EXPERIMENTAL APPROACH

The impact of reducing HCN4 channel function on seizure susceptibility and neuronal network excitability was studied using an HCN4 channel preferring blocker (EC18) and a conditional brain specific HCN4 knockout mouse model.

KEY RESULTS

EC18 (10 mg·kg^-1 ) and brain-specific HCN4 channel knockout reduced seizure susceptibility and proconvulsant-mediated cortical spiking recorded using electrocorticography, with minimal effects on other mouse behaviours. EC18 (10 μM) decreased neuronal network bursting in mouse cortical cultures. Importantly, EC18 was not protective against proconvulsant-mediated seizures in the conditional HCN4 channel knockout mouse and did not reduce bursting behaviour in AAV-HCN4 shRNA infected mouse cortical cultures.

CONCLUSIONS AND IMPLICATIONS

These data suggest the HCN4 channel as a potential pharmacologically relevant target for anti-seizure drugs that is likely to have a low side-effect liability in the CNS.

Using a Multiplex Nucleic Acid in situ Hybridization Technique to Determine HCN4 mRNA Expression in the Adult Rodent Brain

Hyperpolarization-activated cyclic nucleotide-gated (HCN) channels carry a non-selective cationic conductance, I_h, which is important for modulating neuron excitability. Four genes (HCN1-4) encode HCN channels, with each gene having distinct expression and biophysical profiles. Here we use multiplex nucleic acid in situ hybridization to determine HCN4 mRNA expression within the adult mouse brain. We take advantage of this approach to detect HCN4 mRNA simultaneously with either HCN1 or HCN2 mRNA and markers of excitatory (VGlut-positive) and inhibitory (VGat-positive) neurons, which was not previously reported. We have developed a Fiji-based analysis code that enables quantification of mRNA expression within identified cell bodies. The highest HCN4 mRNA expression was found in the habenula (medial and lateral) and the thalamus. HCN4 mRNA was particularly high in the medial habenula with essentially no co-expression of HCN1 or HCN2 mRNA. An absence of I_h-mediated “sag” in neurons recorded from the medial habenula of knockout mice confirmed that HCN4 channels are the predominant subtype in this region. Analysis in the thalamus revealed HCN4 mRNA in VGlut2-positive excitatory neurons that was always co-expressed with HCN2 mRNA. In contrast, HCN4 mRNA was undetectable in the nucleus reticularis. HCN4 mRNA expression was high in a subset of VGat-positive cells in the globus pallidus external. The majority of these neurons co-expressed HCN2 mRNA while a smaller subset also co-expressed HCN1 mRNA. In the striatum, a small subset of large cells which are likely to be giant cholinergic interneurons co-expressed high levels of HCN4 and HCN2 mRNA. The amygdala, cortex and hippocampus expressed low levels of HCN4 mRNA. This study highlights the heterogeneity of HCN4 mRNA expression in the brain and provides a morphological framework on which to better investigate the functional roles of HCN4 channels.