Introduction

Episodic diseases are among the most common to afflict the nervous system. They include migraine headache and epilepsy, which take a great toll on those suffering from these ailments. Genetic contributions to disorders such as epilepsy and migraine are well known, although they are very complex. Understanding the complex genetic milieu of migraine and epilepsy pathogenesis is further confounded by environmental influences that have dramatic effects on the expression of these phenotypes.

One approach to studying complex disorders is through study of a subset of these disorders in which the phenotype is segregating as a Mendelian trait. Such disorders are more amenable to molecular characterization and may provide insights into possible genetic contributions to the complex forms of these disorders.

Episodic phenotypes are common among neurological diseases

There are well-described intermittent disorders of skeletal muscle (periodic paralysis), cerebellar dysfunction (episodic ataxia), and other episodic movement disorders (paroxysmal dyskinesias).¹ Although on the surface, these disorders appear quite different from epilepsy or migraine, they overlap with regard to many of their clinical features. Similar precipitating factors contribute to attacks in all of these disorders. They include stress and fatigue as well as certain dietary factors. Similar responses to various therapeutic manipulations can be seen across these disorders. Finally, some of these disorders demonstrate phenotypes with dysfunction of different parts of the nervous system or peripheral tissues. For example, in Andersen/Tawil syndrome (also known as Andersen's syndrome), hyperexcitability of both skeletal muscle and heart is seen.² Also, in paroxysmal kinesigenic dyskinesia, both an episodic movement disorder and infantile convulsions are recognized as part of the phenotype.³

Mutant ion channels cause intermittent phenotypes in the nervous system

It was less than a dozen years ago that the first voltage-gated ion channel gene mutation was shown to be associated with a phenotype in humans.^4,5 Following this, many sodium channel mutations have been shown to cause at least five different clinical phenotypes. Further work has shown that calcium, chloride, and potassium channel mutations can cause similar muscle diseases.¹ Subsequently, homologs of these channel genes that are expressed in different tissues have been shown to cause various genetic forms of cardiac arrhythmia, epilepsy, and even headache syndromes.

Audiogenic seizures represent a form of reflex epilepsy

Seizures induced by loud sounds have long been recognized in inbred rodent strains. Genetic factors clearly contribute to audiogenic seizure susceptibility, although many such models are multifactoral. Only one model of audiogenic mouse epilepsy has been recognized as a Mendelian trait. This model, first described by Frings and Frings (the Frings mouse) is caused by a gene on mouse chromosome 13.⁶

Reflex epilepsy describes a condition where seizures are provoked habitually by an external stimulus, and sometimes by internal mental processes. Individuals who suffer from reflex epilepsy may have seizures in response to specific stimuli and not suffer spontaneous seizures. In contrast, some reflex seizures may coexist with spontaneously occurring seizures. Reflex epilepsies may manifest as either focal onset or primary generalized seizures, and epileptiform EEG changes are often present.

Audiogenic seizures are recognized in humans but only very rarely. Other types of reflex epilepsy are well known and included seizures induce by flashing lights, auditory, olfactory, or vestibular stimuli. Photosensitive epilepsy is the most common type of reflex epilepsy, occurring when an individual is exposed to visual stimuli (usually flashing lights at a particular frequency). Reflex seizures also may occur in response to more elaborate stimuli or mental processes and include examples such as musicogenic epilepsy or seizures induced by mathematical calculations or reading.

A novel gene implicated in audiogenic seizures in the Frings mice

Study of a large colony of Frings mice led to identification of the causative gene using positional cloning.⁶ Interestingly, the causative gene bore no homology to any known ion channels although a repetitive motif in the encoded protein was homologous to a calcium-binding motif from other proteins.⁷ Once this gene was identified, it was noted that message levels were at extremely low levels in all tissues studied. Thus, while putative alternate transcripts had been recognized, it was difficult to establish which of these were real, since Northern blotting did not yield any measurable message. This problem was exacerbated by the large size of the message (approximately 10 kb). However, one alternate transcript that could be confirmed by RT-PCR from both human and mouse brain RNA led to a frame shift and a truncated protein. The human orthologous gene, hMass1, was mapped to chromosome 5q14, where significant evidence of linkage to febrile seizures (FS) had been noted (FEB4). Screening of hMass1 in individuals from subjects with familial FS revealed a nonsense mutation (S2652X) causing a deletion of the C-terminal 126 amino-acid residues in one family with febrile and afebrile seizures. Thus, a loss of function mutation in hMASS1 might be responsible for the seizure phenotypes in that family.⁸

After we identified the MASS1 gene, another gene was described that had been isolated based on screening by homology with probes from G-protein-coupled receptors.⁹ This gene was named a very large G-protein-coupled receptor (VLGR1) and itself was approximately 10 kb in size. More recently, it has been demonstrated that MASS1 and VLGR1 are two halves of the same gene (McMillan et al¹⁰ and our unpublished data) encoding a protein of the secretin-receptor-like subfamily of GPCRs. The recognition of MASS1 implicated this gene as important in maintaining normal cortical excitability. The G-protein-coupled receptor motif from the VLGR1 gene allows prediction of possible function for this protein in vivo. The full-length transcript now encompasses nearly 20 kb; we now call it "MASSive G-protein-coupled receptor" (MGR1). Although GPCRs are known to function in signaling, much work remains to be done to identify the ligand for this receptor and the role of MGR1 in mammalian brain. The story is further complicated by the fact that the one alternative transcript for which compelling evidence exists has a shifted reading frame leading to a truncated protein that is much shorter than the full-length isoform and does not contain the GPCR motif.

Not all episodic diseases of the nervous system are channelopathies

Based on the precedent of other Mendelian disorders with intermittent phenotypes, ion channels were considered good candidates as the causative gene in audiogenic epilepsy. They are attractive candidates since a widely expressed neuronal channel gene could easily cause a slight alteration of neuronal excitability, which could lead to a phenotype such as epilepsy. However, the identification of MGR1 as an epilepsy gene demonstrates that not all epilepsies will turn out to be caused by ion channels.

Given the phenotype of audiogenic seizures and the pathway of auditory inputs to the brain, we hypothesize that MGR1 may not function in determining primary membrane excitability. Rather, MGR1 may be involved in signal integration of brainstem inputs and outputs. An alternate and very interesting possibility is that MGR1 is important in development of nervous system circuitry and that abnormal 'wiring' leads to predisposition to hyperexcitability.

Another seizure disorder, autosomal–dominant partial epilepsy with auditory features (ADPEAF), is a rare form of idiopathic lateral temporal lobe epilepsy characterized by partial seizures with auditory disturbances. Recently, this disorder was shown to be caused by mutations in the leucine–rich, glioma–inactivated 1 gene (LGII) in five ADPEAF families.¹¹

That MASS1 and LGI1 turned out not to be ion channel genes makes the molecular characterization more difficult. Both were identified through a positional cloning approach. While quite laborious, it is exciting to identify novel epilepsy genes. All current anticonvulsants modulate ion channel function. Novel targets for development of new anticonvulsants that work via different pathways than currently available antiseizure drugs may provide new opportunities for treating epilepsy.

Analysis of the sequences revealed an interesting common feature. Based on predicted architectural and structural features, this seven-fold repeated 44-residue motif suggested that it belongs to a group of protein interaction domains containing a seven bladed –propeller fold.¹² Kay Hofmann's group termed these EAR domains for epilepsy associated repeats. This is an excellent example of the power of bioinformatics and the complimentary role this discipline brings to biology and genetic worlds.

Identification of the MASS1 gene has opened up exciting possibilities into the pathogenesis of audiogenic epilepsy. However, many questions remain before the role of a mutation in this G-protein-coupled receptor with EAR domains might ultimately lead to an epilepsy phenotype.

References

1. Jen J, Ptácek LJ. In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds). Metabolic and Molecular Bases of Inherited Disease, 8th edn. McGraw-Hill: New York, 2001, pp. 5223–5238.
2. Sansone V et al. Ann Neurol 1997; 42: 305–312. | Article | PubMed | ISI | ChemPort |
3. Swoboda KJ et al. Neurol 2000; 55: 224–230.
4. Ptácek LJ et al. Cell 1991; 67: 1021–1027. | Article | PubMed | ISI | ChemPort |
5. Rojas CV et al. Nature 1991; 354: 387–389. | Article | PubMed | ISI | ChemPort |
6. Skradski S et al. Neuron 2001; 31: 537–544. | Article | PubMed | ISI | ChemPort |
7. Schwarz EM, Benzer S. Proc Natl Acad Sci USA 1997; 94: 10249–10254. | Article | PubMed | ChemPort |
8. Nakayama J et al. Ann Neurol 2002; 52: 654–657. | Article | PubMed | ISI | ChemPort |
9. Nikkila H et al. Mol Endocrinol 2000; 14: 1351–1364.
10. McMillan DR et al. J Biol Chem 2002; 277:785–792.
11. Kalachikov S et al. Nat Genet 2002; 30: 335–341. | Article | PubMed | ISI |
12. Scheel H et al. Hum Mol Genet 2002; 11: 1757–1762. | Article | PubMed | ISI | ChemPort |