Investigations of Epilepsy with a Mutant Animal (EL Mouse) Model


  • Jiro Suzuki

    1. Organized Center of Clinical Medicine, Department of Health and Welfare, International University of Health and Welfare, Ohtawara, Tochigi, Japan
    Search for more papers by this author

Address correspondence and reprint requests to Dr. J. Suzuki at Organized Center of Clinical Medicine, Department of Health and Welfare, International University of Health and Welfare, Sannou Medical Plaza, 8-5-35, Akasaka, Minato-ku, Tokyo, 107-0052, Japan. E-mail:


Summary: Purpose: The EL mouse model, with running fits and convulsions, has been useful for studying mechanisms of epilepsy, epileptogenesis, and ictogenesis.

Methods: The history of this model and recent key findings are described.

Results: Epileptogenesis has a hereditary component, and the genes responsible for it are presumed to be multiple. A seizure of an EL animal is precipitated by a rapid accelerating movement. Seizures develop in conjunction with aging of the mouse, repetition of stimuli, and with seizures themselves. This phenomenon represents a type of abnormal plasticity, which underlies epileptogenesis. The paroxysmal discharges in EL begin at parietal cortex, propagate to hippocampus, and then to the entire brain. Manifestation of a seizure requires a combination of several brain regions, termed the “focus complex.” A small but significant disorganization of hippocampal cytoarchitecture occurs, leading to higher excitability. Abnormally low GABAergic function in the parietal cortex and hippocampus develops with maturation of the EL mouse and with repetition of stimuli and seizures. Low superoxide dysmutase activity and abnormal eNOS function in hippocampus or parietal cortex may relate to epileptogenesis of EL. After seizures, immediate early genes (e.g., c-fos and zif) expression and DNA fragmentation are observed, which play important roles in ictogenesis and epileptogenesis. All of these phenomena follow a process of development similar to that of the seizure itself.

Conclusions: Insights from the EL model suggest that epileptogenesis and ictogenesis in epilepsy can be viewed in terms of genetic predisposition and a new concept of abnormal plasticity.

Investigation of the mechanisms of epilepsy using a mutant animal was initiated by Studentsov in rats with audiogenic seizures in the 1920s. Use of mutant animal models has been popular since the 1950s. From 1960 to 1970, epileptic seizures were observed in baboons and mongolian gerbils, and in many kinds of neurological mutants in mice. Some of these mutant mouse models can be studied to great advantage in experimental epileptology.


In 1954, Imaizumi et al. (1) observed mice with hydrocephalus that exhibited convulsive seizures after being tossed into the air. They inbred susceptible mice in order to produce a strain with epileptiform convulsive seizures. This strain was reported as the ep mouse in 1959. Internationally, the mutant animal was registered as El of F25 dominant in 1964 (2). Subsequently, a number of biochemical research reports on the ep or El mouse were published with the title “epilepsy-like convulsions.” We (3) established the El mouse as having epilepsy by recording paroxysmal discharges of the EEG during wild running and tonic–clonic convulsive seizures in 1976. The first symposium on this mouse was held at Tokyo in 1992, where the name of the animal was changed from El to EL.


Seizures of EL are evoked by natural sensory stimulation (e.g., tossing the mouse up into the air, or subjecting it to a rapid acceleration). Effective stimulation acts through the proprioceptive neural pathway (not via the vestibular pathway) (unpublished data). Although this seizure susceptibility is genetically predisposed, repetition of such stimuli is necessary to provoke a seizure. To display the predisposition to seizures, from age 3 weeks, an EL is given provoking stimulation once a week, but no seizure is observed until age 6 weeks. Some ELs exhibit abortive seizures at a younger age. The incidence of seizures increases with repetition of stimulation and prevalence of seizures after stimulation eventually attains 100% (4).

After a long period without stimulation, the seizure threshold of an EL with a fully developed seizure tendency is raised, or extinction of seizures is observed. After stimulation is resumed and repeated, as previously, seizures are induced again in almost all animals. However, after a 14-week absence of stimulation, no seizure occurs at the first resumption of stimulation, but the reappearance of seizures and the lowering of seizure threshold redevelops more rapid than in naïve animals. Occurrence of seizures in the EL requires reinforcement stimulation and is affected by certain types of experience or memory of antecedent stimuli or seizures. We called these phenomena abnormal plasticities in occurrence of a seizure in an EL (4).


The electrical activities of various brain regions during seizures were recorded via depth electrodes and a special preamplifier (5). The first spike of paroxysmal activities starts at the right parietal cortex (PCx) with some initial discharges. Then 0.2 s later, in the left PCx, the next spike follows. The paroxysmal discharges propagate to the hippocampus (HIP), frontal cortex, striatum, substantia nigra, and other areas and cause running fits or tonic convulsions. The HIP, which receives propagated discharges, initiates larger spikes that appear to drive manifest convulsions. The PCx is an “ignition starter,” and the HIP may be a “booster apparatus.” This phenomenon implies that the clinical seizures of the EL should be precipitated by discharges of “the focus complex” composed of the PCx, HIP, and other regions.

Paroxysmal neural activities with seizures are reflected by local metabolic activities. Local cerebral glucose use (LCGU) examined with 2-deoxy-d-glucose (2DG) (6) during an entire seizure and recovery period in an EL significantly increases in the PCx, dorsal HIP, dentate gyrus, ventral thalamus, and cerebellar nuclei (7,8). Using nicotinamide adenine dinucleotide phosphate (NADP) cycling, DG and DG-6P concentrations were measured over a very short term. The concentration of DG-6P increases rapidly at 60–90 s after a seizure has started in the PCx, whereas, in the HIP, the increase of DG-6P is observed 90–120 s later (9).

The γ-aminobutyric acid (GABA)ergic function in cortex of EL was analyzed with an ultramicro-quantitative method using the reaction of cycling of GABA conversion to NADPH (9,10). In the PCx of a control mouse strain, designated ddY, the GABA concentration level gradually increased along the rostrocaudal axis. However, that of an EL showed a remarkably sharp decrease over an approximate 100 × 100-μm width, a “pocket-like” drop. Glutamic acid decarboxylase (GAD) activity levels of control and EL animals paralleled those of GABA. In the fourth internal granular layer in the pocket-like drop region of PCx of an EL brain, both GABA concentrations and GAD activities were very low. By contrast, the GABA concentration and GAD activity in the HIP of an EL brain were twice that in region CA1, but very low in region CA3, compared with control levels.

The abnormal GABAergic function of the EL brain becomes more significant as animals mature and have repeated seizures. At 5 weeks old, the abnormal pocket-like drop is small, but it becomes very prominent at age 30 weeks.


After seizures of an EL, immediate early genes (IEGs, such as c-fos and zif) demonstrate expression in brain with in situ hybridization analysis (10–12). This IEG expression relates to the seizure history, seizure threshold, and age. In conjunction with development of an animal and repetition of seizures, the zif expression site shifts from the pyriform and entorhinal cortices and the dentate gyrus to region CA1. In the PCx, zif expression initially is localized, but later extends diffusely after repetition of seizures. IEG expression manifests continuously in the interictal period, after the seizure threshold becomes very low. These IEG expression sites in an EL brain are almost identical to the regions of the abnormal GABAergic system.

Recent investigation of free radicals suggests that their activities may induce brain pathology by damaging DNA in neurons. The amount of superoxide dismutase (SOD), scavenger of active oxidants, was assayed in the brains of an EL and control ddY (13). In the HIP and PCx of EL brain, the amount of CU/Zn-SOD, which is relatively more than that of control, decreases with development and repetition of seizures. This suggests higher amounts or excess activities of free radicals in the EL brain.

To understand the role of nitric oxide (NO) in the regulation of seizures, three isoforms of nitric oxide synthase (NOS) were assayed by an immunoblot analysis of HIP tissues (14). nNOS is the major component and iNOS the minor component, but in excess amounts in an EL. However, eNOS, although in very small amounts, appears to be responsible for NO that mediates increase in LCGU during focal seizures.

The target of excess free radicals may be DNA, primarily resulting in DNA fragmentation. DNA fragmentation without cell loss is found in the HIP region CA1 and PCx of the EL brain (10). Continuously abnormal GABAergic function, IEG expression, and excess activities of free radicals are involved in epileptogenesis of EL. Conversely, transient IEG expression, an excess of iNOS, and a very small amount of eNOS are associated with ictogenesis of an EL seizure (14).


Stimulation of the perforant path of the hippocampus evokes potentials in the dentate gyrus. A laminar profile of the evoked dentate potentials shows a greater spatial extent and higher amplitude of excitatory postsynaptic potentials (EPSPs) and population spikes in the ELs with repeated seizures than either in the ELs with no seizures or in controls (15).

The laminar structure of the pyramidal cell layer in regions CA1 and CA2 is more loosely arranged, particularly at the boundary zone of the strata, in the EL, compared to the ddY. The cells of EL seem larger and less dense than do those of ddY. The granule cell layer in the dentate shows similar features in the EL, compared to ddY (16), whereas no notably abnormal finding is found in the PCx.

To reiterate, abnormal GABA function, excess free oxidants, abnormal composition of NO isozymes and IEG expression after a seizure and an interictal period, which are observed in the HIP as well as in the PCx, play important roles in establishing epileptogenesis. Ictogenesis in EL is mediated by abnormal plasticity (Y.L. Murashima, M. Yoshii, and J. Suzuki, unpublished observations).


Seizure tendency in EL is inherited phenomenologically as an autosomal dominant trait. In 1991, Seyfried et al. (17) showed the inheritance of the EL to be polygenic (i.e., inheritance involves at least two genes); EL 1 is localized at the region distal to the centromere on chromosome 9, and EL 2, at chromosome 2. Furthermore, Seyfried and associates (18) reported that various environmental risk factors influence the development of seizure susceptibility in the EL.


The term neural plasticity refers to the capacity of the nervous system to exhibit structural and functional adaptations to impinging stimuli (19). Many reports indicate that altered gene expression contributes to the processes that underlie neural plasticity. Both c-fos and zif are mRNAs that contribute to the long-term synaptic plasticity process. In a seizure model by Curran et al. (20) and a kindling model by Dragunow et al. (21), c-fos mRNA changed in a dramatic manner. Prolonged administration of cocaine or amphetamine can produce characteristic psychotic manifestations, in which similar processes of IEG expression and transcription to DNA precede neuronal alterations.

Some responses are provoked in neurons according to a special predisposition or vulnerability, which originally is determined by genetics (e.g., the EL mouse). In such responses, gene expression is induced to produce new peptides or proteins. By nature, a living organism responds to various input stimulations, whether they are noxious or not. If an organism has genetic abnormalities relating to receptive or other functions, a response that otherwise would extinguish will instead alter neuronal function and lead to abnormal symptoms. The mechanism that should underlie these phenomena is termed abnormalplasticity, which can be added to the established processes of pathogenesis.


The EL mouse is an excellent model of genetic and secondarily generalized epilepsy. The running fit and/or convulsive seizure tendencies are inherited as autosomal dominant traits. The genes locate in multiple loci, and environmental risk factors affect the occurrence of seizures. The paroxysmal discharges initiate at the PCx and propagate to the HIP and the entire brain. Manifestation of seizures requires a combination of several brain regions, the “focus complex.” A small but significant disarrangement of cytoarchitecture of the HIP occurs, associated with increased excitability.

Expression of seizures is a function of mouse maturation, repetition of stimuli and of seizures themselves. This phenomenon, abnormal plasticity, underlies epileptogenesis. Abnormally low GABAergic function, which exists in the PCx and HIP, develops with growth of a mouse and repetition of stimuli and seizures. A low activity of SOD and abnormal eNOS function, which are observed in the HIP or the PCx, may relate to epileptogenesis in EL. After seizures, IEGs (e.g., c-fos and zif) expression and DNA fragmentation are observed, which play an important role in ictogenesis and epileptogenesis. Abnormal plasticity may be a new concept related to a process underlying pathogenesis of neuropsychiatric disorders.