Some of the most common diseases in humans occur intermittently in people who are otherwise healthy and active. Such disorders include migraine headache, epilepsy, and cardiac arrhythmias. All of these diseases have prominent genetic components. Difficulty in understanding disorders like epilepsy and migraine arises from the complexity of the clinical phenotypes and the genetic heterogeneity that is certain to exist. Therefore, work in our lab has aimed at understanding the pathogenesis of rare monogenic disorders that are similar in their episodic nature.
We have identified and studied five ion channel genes that, when mutated, cause a group of familial muscle diseases that share many similarities with other episodic disorders, such as headache and epilepsy. They include genes that encode sodium, calcium, potassium, and chloride channels—and are thus termed "channelopathies." More recently, we have extended this work to episodic disorders of the brain. In some cases, study of the effects of mutations in these genes and correlation of gene mutations with the patients' phenotypes have led to improvements in physicians' abilities to diagnose and treat patients.
Translating Findings from Rare Families to Common and Complex Disorders
Thyrotoxic hypokalemic periodic paralysis (TPP) is characterized by acute attacks of weakness, hypokalemia, and thyrotoxicosis of various etiologies. These transient attacks resemble those of patients with familial hypokalemic periodic paralysis, except that it is a sporadic condition. The attacks are prevented by treating the underlying hyperthyroidism. Worldwide, TPP is at least 10 times more common than all the familial periodic paralyses together and affects ~10 percent of all thyrotoxic Chinese men. Because of the phenotypic similarity of these conditions, we hypothesized that TPP might also result from mutations in ion channels that were uncovered only with elevated thyroid hormone levels. While sequencing candidate genes, we identified a gene that encodes a novel inwardly rectifying potassium (Kir) channel, Kir2.6. It is highly similar to Kir2.2, expressed in skeletal muscle, and transcriptionally regulated by thyroid hormone. Expression of Kir2.6 in mammalian cells revealed normal Kir currents in whole cell and single channel recordings. Kir2.6 mutations were present in 33 percent of the unrelated TPP patients in our collection. Some of these mutations alter a variety of Kir2.6 properties, all of them altering muscle membrane excitability and leading to paralysis.
Not All Episodic Diseases Result from Ion Channel Mutations
Ion channel disorders are now recognized not only in muscle but also in heart and brain. Interestingly, there are still other Mendelian episodic diseases that have not been characterized at a molecular level.
The familial paroxysmal dyskinesias are a group of overlapping clinical phenotypes that segregate in families. On the surface, the paroxysmal dyskinesias are quite different from other episodic disorders. However, they also share many similarities, such as the precipitants, which can result in attacks and the response to similar medications like anti-seizure drugs. We have cloned a gene for paroxysmal nonkinesigenic dyskinesia (PNKD). Rather than being a channel, the PNKD gene encodes a novel protein with homology to enzymes important for glutathione-coupled detoxification in a stress response pathway. We have shown that this novel protein is localized to the synapse and regulates exocytosis. We have now generated mice carrying the mutant human gene. These mice manifest a disorder very similar to the human disease and are enabling work to elucidate the role of this protein in brain and dysfunction of the mutant protein in PNKD. Since ion channels have been the major target for development of migraine and epilepsy drugs, non–channel proteins encoded by these genes, such as the one causing PNKD, may provide novel targets for development of a new class of treatments for episodic disorders.
We have also localized genes causing a familial, adult-onset, myoclonic epilepsy to chromosome 8 and an epilepsy and movement disorder to chromosome 16 and continue efforts to identify the causative genes and mutations. In both cases, we've ruled out ion channel gene mutations. Thus, identifying these genes and studying the encoded proteins will lead to novel insights into the electrical excitability of neurons and the brain.
Complex Episodic Disorders
Identifying genes that cause complex, polygenic disorders such as epilepsy and migraine is much more difficult than cloning genes that cause Mendelian diseases. Epilepsy and migraine share many features with the channelopathies described here. These data suggest that there may be some similarities between the physiological basis of migraine, epilepsy, and rare, monogenic, episodic nervous system disorders. Identification of genes and mutations causing rare and strongly genetic forms of these diseases continues to allow testing of hypotheses regarding 'risk' alleles that contribute to common forms of episodic disorders.
Human Sleep Behavior
Large variations exist among all people with regard to preferred sleep and wake times, in a spectrum from morning larks to night owls. Genetic clues from Mendelian sleep variants may shed light on common genetic variants that contribute to human sleep preferences. Advanced sleep-phase syndrome is a common phenomenon in aging individuals (ASPS of aging). In this syndrome, people fall asleep and wake earlier than desired and earlier than they did when they were younger. A third of 65-year-olds wake significantly earlier than they did as young adults, but the basis of the ASPS of aging is not known.
We have identified a Mendelian circadian rhythm trait in humans resulting from a short circadian period. This syndrome, familial advanced sleep-phase syndrome (FASPS) is similar to the ASPS of aging except that it occurs in the young, is transmitted as a Mendelian trait, and is a more dramatic advance of the sleep time preference (extreme "morning larks"). In collaboration with Ying-Hui Fu (University of California, San Francisco), we have mapped and cloned five genes responsible for this trait in FASPS pedigrees. One gene, hPER2, is a homolog of the Drosophila period gene. The mutation occurs at a serine residue (S662) in a location that we have shown to be within the binding site of casein kinase I (CKI), an enzyme that phosphorylates PER2. A second gene encoding casein kinase Iδ (CKIδ) also harbors an FASPS mutation. This mutation of CKIδ leads to hypophosphorylation of PER2 in vitro. Moreover, phosphorylation of PER2 S662 leads to a hierarchical phosphorylation cascade of downstream serines by CKI. Further work is directed at identifying other substrates and establishing which of them are important in circadian regulation.
Mouse models of the Per2 and CKIδ mutations have been generated and show phenotypes identical to that seen in human FASPS. These mice, which have allowed in vivo study of the functional consequences of these mutations at a basic molecular level, have revealed some surprises and led us to modify existing models of circadian function.
In collaboration with Christopher Jones (University of Utah), we have collected a large number of FASPS families that are being investigated for mutations in hPER2, CKIδ, and other circadian rhythm candidate genes. We've also now identified families with the opposite phenotype—familial delayed sleep phase (FDSPS—extreme "night owls"). Mutations in two circadian genes have been identified.
Although we spend about one-third of our lives sleeping, almost nothing is known about what sleep is and why we need it. It is well appreciated that altered sleep quantity (deprivation) and altered phase can increase risk for many disorders, including cancer, metabolic syndrome, heart disease, and others—but the reason for this increased risk is not known.
We have recently described a phenotype in which affected individuals have a lifelong ability to sleep much less than average and still feel well rested and energetic. We call this familial natural short sleep (FNSS). Circadian clock and sleep homeostasis are known to be distinct. At the same time, these two processes also have some overlap. Thus, we searched for genetic variants in circadian rhythm genes. Mutations in two circadian genes have been identified and further characterized. Remarkably, 90 percent of our FNSS families do not have mutations in circadian genes. These families are a remarkable resource for identifying novel genes critical for insights into novel pathways necessary for sleep regulation and may shed light on mechanisms by which altered sleep contributes to increased risk for other diseases.
In summary, genetic approaches to neurologic diseases and behavioral phenotypes have made great progress in the past decade. Such discoveries are leading to insights not only into disease pathophysiology but also into normal function of the nervous system and normal variation in the population. More recently, the genetic contributions to human behavior have begun to be elucidated. Ultimately, such knowledge will lead to better diagnosis and treatment of patients with neurological diseases.
Grants from the National Institutes of Health provided partial support for these projects.
As of March 23, 2016