Lateral temporal lobe epilepsies: Clinical and genetic features

Authors


Address correspondence to Roberto Michelucci, Department of Neurosciences, Division of Neurology, Bellaria Hospital, via Altura, 3, 40139 Bologna, Italy. E-mail: Roberto.Michelucci@ausl.bo.it

Summary

Lateral temporal epilepsies are still a poorly studied group of conditions, covering lesional and nonlesional cases. Within nonlesional cases, autosomal dominant lateral temporal epilepsy (ADLTE) is a well-defined, albeit rare, condition characterized by onset in adolescence or early adulthood of lateral temporal seizures with prominent auditory auras sometimes triggered by external noises, normal conventional magnetic resonance imaging (MRI), good response to antiepileptic treatment, and overall benign outcome. The same phenotype is shared by sporadic and familial cases with complex inheritance. Mutations in the LGI1 gene are found in about 50% of ADLTE families and 2% of sporadic cases. LGI1 shows no homology with known ion channel genes. Recent findings suggest that LGI1 may exert multiple functions, but it is not known which of them is actually related to lateral temporal epilepsy.

Lateral temporal lobe epilepsies (LTEs) are currently distinguished from mesial temporal lobe epilepsies, particularly because of specific seizure characteristics—such as auditory auras, aphasic seizures, and high propensity to generalize—and clinical and etiological context, lesional cases being typically uncommon (Florindo et al., 2006). Moreover, patients with LTE account for only a minority (about 10%) of all temporal epilepsies (Williamson & Engel, 2008).

A number of families and sporadic cases with nonlesional LTE have been reported in past years, the former under the terms autosomal dominant lateral temporal epilepsy (ADLTE) or autosomal dominant partial epilepsy with auditory features (ADPEAF), and the latter as idiopathic partial epilepsy with auditory features (IPEAF). Clinically, both conditions are characterized by focal seizures with prominent ictal auditory phenomena, negative magnetic resonance imaging (MRI) findings, and relatively benign evolution (Ottman et al., 1995; Michelucci et al., 2003; Bisulli et al., 2004a).

Mutations in the leucine-rich, glioma-inactivated 1 (LGI1) gene have been found in about 50% of ADLTE families (Michelucci et al., 2003; Ottman et al., 2004) and in about 2% of sporadic IPEAF cases (Bisulli et al., 2004b;Michelucci et al., 2007). The LGI1 gene is mainly expressed in brain tissues and shows no homology with known ion channel genes. It encodes a protein containing two structural domains: four leucine-rich repeats (LRR) (Kobe & Deisenhofer, 1994) in the N-terminal half, and seven epitempin (EPTP) repeats (pfam # PF03736) in the C-terminal portion of the protein. Both domains likely mediate protein–protein interactions. Here we summarize the clinical features of familial and nonfamilial patients with LTE and illustrate the nature and the possible functional consequences of LGI1 mutations.

Clinical Features

ADLTE/ADPEAF is a rare familial condition so far reported in Europe, the United States, Australia, and Japan. The real prevalence of ADLTE is unknown, but it may account for about 19% of familial idiopathic focal epilepsies (Ottman, personal communication). Since the first description of the syndrome (Ottman et al., 1995), 35 families have been published covering a total of almost 200 patients. The syndrome segregates with an autosomal dominant inheritance pattern with incomplete penetrance (70–80%). The families are selected by demonstrating the existence of at least two cases over two generations with unprovoked focal or secondarily generalized seizures, the symptoms of which suggest a lateral temporal lobe onset.

The age of onset ranges between 1 and 60 years with a mean of 18 years. The ictal semiology includes focal seizures (usually elementary) and secondarily generalized tonic–clonic seizures. Focal seizures are characterized by auditory auras in 64% of the cases. Other less frequent auras include complex visual (17%), psychic (16%), autonomic (12%), vertiginous (9%), and other sensory (13%) symptoms; aphasic seizures associated or not with auditory phenomena are reported in 17% of the overall ADLTE patients and were the only clinical symptom in at least two pedigrees (Michelucci et al., 2003). Tonic–clonic seizures are common, occurring in 90% of cases, both during wakefulness and sleep, and may unmask an otherwise undiagnosed history of elementary focal seizures with auditory symptoms.

Auditory auras are described as elementary and simple sounds (such as humming, buzzing, and ringing) in most cases (74%) or as complex hallucinations (i.e., music, voices) in a minority of patients (11%). Sometimes the aura is characterized by a distortion of sounds (becoming louder and louder or suddenly low) (28%). Neurophysiologic studies in ADLTE have mainly focused on interictal electroencephalography (EEG), which shows temporal abnormalities (described as mild slow/sharp waves) in 47% of the patients. Interestingly a clear left predominance of the abnormalities has been emphasized in some families by interictal and ictal EEG documentation, asymmetry of long-latency auditory evoked potentials (Brodtkorb et al., 2005), and impairment of dichotic listening performance (Pisano et al., 2005). MRI studies have shown no abnormalities in most pedigrees. Kobayashi et al. (2003) described a Brazilian ADLTE family associated with LGI1 mutation in which most affected individuals showed left temporal lobe “malformations,” but this neuroradiologic trait did not segregate entirely with the genotype. Tessa et al. (2007), by using nonconventional MRI techniques (voxel-based analysis of diffusion tensor MR images) in eight ADLTE mutated patients, demonstrated a cluster of fractional anisotropy in the left lateral temporal cortex, suggesting a malformative origin of the abnormality.

Genetic studies have revealed mutations of LGI1 in about 50% of the families, providing evidence of genetic heterogeneity. Detailed analysis of European and U.S. families with and without LGI1 mutations showed no phenotypic differences between the pedigrees; the only different finding, reported by Ottman et al. (2004) in nonmutated families, was a higher frequency of seizures with autonomic features and a lower frequency of auditory auras compared with mutated families.

Following the extensive search for new families, a number of pedigrees have been identified, with only one patient showing auditory auras in a context of complex or multifactorial inheritance pattern. These cases are invariably found without any LGI1 mutation and should, therefore, be kept apart from ADLTE and better referred to as atypical familial LTE.

Sporadic, nonfamilial and apparently nonsymptomatic cases with auditory seizures have been collected in past years in order to discover a possible role of LGI1 in these more common cases of LTE. In 2004, Bisulli and collaborators reported 53 unrelated cases with a homogeneous clinical picture characterized by negative family history, a mean age of onset of 19 years, focal seizures with prominent auditory auras, a high rate (79%) of secondary generalized tonic–clonic seizures, low seizure frequency, good response to antiepileptic treatment, tendency to recurrence after drug withdrawal, unrevealing EEG, and normal MRI. Comparison with familial cases showed no significant clinical difference and supported the existence of distinct entity (referred to as IPEAF) with a phenotype closely resembling ADLTE but without family history. To date, de novo LGI1 mutations have been found in two IPEAF cases (see subsequent text).

Genetic Features

Twenty-three ADLTE-causing LGI1 mutations have been reported, nine of which result in protein truncation or deletion and 14 in single amino acid substitutions (Gu et al., 2005; Chabrol et al., 2007; Striano et al., 2008). Both truncating and missense mutations are evenly scattered along the gene. Overall, no major clinical differences have so far been found between families with either truncating or missense mutations, and with point mutations involving the LRR or EPTP domain. Two de novo LGI1 mutations have been identified in a collection of 104 sporadic IPEAF cases (Bisulli et al., 2004b; Michelucci et al., 2007), one of which, the nonsense mutation c.1420C>T, has also been found in an ADLTE family, whereas the other, R136W, is associated with a peculiar form of telephone-induced auditory partial epilepsy.

The mechanism by which LGI1 mutations determine epilepsy remains unclear. Cell transfection experiments have shown that the wild type Lgi1 protein produced in vitro is secreted (Senechal et al., 2005). On the contrary, all the epilepsy-associated mutant proteins tested so far, either truncated or carrying single amino acid substitutions, are retained within the cells (Senechal et al., 2005; Chabrol et al., 2007; Striano et al., 2008). Interestingly, point mutations occurring in the LRR or EPTP domain have the same negative effect on protein secretion, suggesting that both domains are necessary for secretion. Based on these experimental data, it is currently believed that Lgi1 may function as a ligand in the extracellular or synaptic environment. In line with this hypothesis, the Lgi1 protein has been shown to bind to the postsynaptic receptor ADAM22 and to affect the glutamate-α-amino-3-hydroxyl-5-methyl-4-isoxazole propionate acid neurotransmission (Fukata et al., 2006). On the other hand, other experiments suggest that Lgi1 may act as an ancillary subunit of the presynaptic Kv1 potassium channels implicated in the regulation of channel inactivation (Schulte et al., 2006), a putative function not necessarily requiring secretion of the Lgi1 protein.

An alternative functional hypothesis implicates LGI1 in central nervous system (CNS) development. This hypothesis is supported by: (1) the structural homology of the Lgi1 LRR domain with that of other LRR proteins essential for the CNS development, such as slit, toll (Kobe & Deisenhofer, 1994); (2) experimental evidence showing that LGI1 is involved in the control of proliferation and survival of neuroblastoma cell lines, suggesting a similar function in neuronal development (Gabellini et al., 2006); and (3) the subtle structural anomalies detected in the left lateral temporal cortex of LGI1-mutated epileptic patients by voxel-based diffusion tensor MR analysis, which may well underlie partial epilepsy in these patients (Tessa et al., 2007).

Conclusions

There is a spectrum of nonlesional clinical entities within the LTEs, ranging from familial cases (with either strict autosomal dominant—ADLTE—or complex inheritance) to sporadic nonfamilial patients. These entities share a common phenotype characterized by partial seizures with prominent auditory auras but seem to have a different genetic basis, with only ADLTE being substantially related to LGI1 mutations. The LGI1 mutations so far identified appear to be associated with a rather homogeneous clinical phenotype. Analysis of a larger set of mutations together with a detailed description of the clinical phenotype of ADTLE will provide insights into genotype–phenotype correlations.

Acknowledgments

This work was supported by the Genetic Commission of the Italian League Against Epilepsy.

We have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Disclosure: The authors have no conflicts of interest to declare.

Ancillary