Precision physiology and rescue of brain ion channel disorders

Abstract

Ion channel genes, originally implicated in inherited excitability disorders of muscle and heart, have captured a major role in the molecular diagnosis of central nervous system disease. Their arrival is heralded by neurologists confounded by a broad phenotypic spectrum of early-onset epilepsy, autism, and cognitive impairment with few effective treatments. As detection of rare structural variants in channel subunit proteins becomes routine, it is apparent that primary sequence alone cannot reliably predict clinical severity or pinpoint a therapeutic solution. Future gains in the clinical utility of variants as biomarkers integral to clinical decision making and drug discovery depend on our ability to unravel complex developmental relationships bridging single ion channel structure and human physiology.

Figure 1.

Pore and non-pore impact of ion channels in developing brain. Ion channelopathy is cell type and network specific and may produce reversible and irreversible alterations in brain function at multiple stages of development.

Figure 2.

Match and mismatch of mutant channel expression in channelopathy phenotypes. (A–D) In situ hybridization of ion channel mRNA transcripts show concordant (A and B) and discordant (C and D) expression in brain circuits mediating mutant channel phenotypes. (A) Expression of Kcnma1 subunits linked to human thalamocortical absence epilepsy phenotype show strong matching expression in cortical-thalamic (TCR, nRT nuclei) circuit. (B) Mutation of the related β subunit Kcnmb4 in mouse lacks significant thalamic expression (circle); without it, the mutant mouse shows a hippocampal epilepsy phenotype of temporal lobe epilepsy instead. (C) Diffuse expression of Cacna1a transcripts broadly overlaps with pathogenic thalamocortical absence seizure pathway and with cerebellar ataxia, but does not clinically affect most other synaptic pathways. (D) Mutation of the related β subunit Cacnb4 shows identical seizure/ataxia phenotype as Cacna1a mutant mice, but somewhat better defines thalamocortical disease circuitry.

Figure 3.

Genetic isolation of critical epilepsy circuit underlying widespread calcium channelopathy. (A and B) Cre-driven selective ablation of Cacna1a in layer 6 corticothalamic projection neurons only delimits critical synapses sufficient for appearance of thalamocortical seizures, shown in C. (D) Defective glutamate release (N-type, rather than P/Q type) at these thalamic synapses induces postsynaptic enhancement of low threshold T-type calcium current and rebound bursting in downstream thalamic cells. Remodeling caused by loss of developmental homeostasis at this single class of synapses results in diffusely hypersynchronized cortical networks (reprinted from Bomben et al., 2016).

Figure 4.

Maturation of ion channel expression and firing properties in developing interneurons. Channel subunit switching in a fast spiking neocortical inhibitory interneuron demonstrated by laser capture of parvalbumin + cells at differing ages and analyzed by microarray for sodium, potassium, and calcium channel subunit transcripts. Many subunits show clear reversal of channel transcript pattern in second postnatal week (reprinted from Okaty et al., 2009).

Figure 5.

Human ion channel variant complexity requires massive simulation to solve complex personal excitability profiles. (Left) Unexpected complexity of novel nonsynonymous single nucleotide variants detected among 237 channel subunits sequenced in two individuals with epilepsy (upper plots, affected 1 and 2) and without (lower, control 1 and 2). Overall, the numerical burden of channel variants did not significantly differ between these groups (epilepsy adults n = 152, neurological normal adults n = 139), indicating that pattern rather than load is a major contributor to phenotype. (Right) Computer simulation of a single hippocampal neuron firing pattern when current amplitudes are varied in a simple “two hit” model of a digenic mutation interactions between Na_v/Ca_v (A) and Na_v/K_v (B). In the future, personalized models of complex masking and degenerate current/firing pattern outcomes can be systematically tested in large networks incorporating an individual’s full compound variant profile (“channotype”) and correlated with real-life sensitivity to ion channel-based therapies (reprinted from Klassen et al., 2011).

Figure 6.

Secondary channelopathy resulting from mutation of nonsubunit interacting gene. ADAM11 is critical for retention of K_v1.1/K_v1.2/K_vβ2 subunits at the presynaptic terminal (arrows) and "pinceaux" (arrowheads) of basket cells onto Purkinje cells (asterisks). Tethered channel subunit proteins such as ADAM11, a member of the large membrane disintegrin and metalloproteinase family, play a key role in selective targeting and retention of K_v1.x channels to wild-type presynaptic terminals, as shown by their absence at cerebellar inhibitory basket cell terminals (yellow boxes) in the epileptic ADAM11^Δ12–18 (a truncation removing the domain containing the integrin-binding site) mutant mouse. HCN1, another channel located at the basket cell presynaptic terminal, is not altered. Bars: (left) 50 μm; (insets) 10 μm. (Bottom) The same heteromeric K1.x channel α subunits are spared at mutant peripheral nodes of Ranvier, which depend on ADAM22 (modified with permission from Kole et al., 2015).

Figure 7.

Transcompartmental rescue of K_v1.1 potassium channel null mutation by opening neighboring KCNQ2 current. (A) Genetic deletion of repolarizing K_v1.1 channel normally residing in flanking paranodal regions provokes generation of spontaneous and stimulated ectopic impulses recorded in isolated mouse vagal axons (B and C). (D) Hyperexcitability can be restored to wild type level by opening KCNQ2 channels in nodal membrane with flupirtine. The identification of currents that reside in the vicinity of a mutated channel may predict useful therapeutic targets (modified with permission from Glasscock et al., 2012).