Microscopic brain structure revisited in genetic epilepsy.

There are usually no structural changes detectable on an individual MRI scan of a patient with idiopathic/genetic generalized epilepsy (IGE/GGE) or simple febrile seizures. This observation is also used as a diagnostic criterion for this type of epilepsy, and more broadly also for other types of genetic epilepsies. However, slight morphologic changes can be detected as group effects in comparison to normal controls, using, for example, postmortem histologic investigations and in vivo MRI scans of human brains. In this issue of Neurology®, Wimmer et al. add a new aspect to this long-term ongoing discussion by reporting changes in the number of GABAergic inhibitory neurons in different cortical layers in a mouse model of genetic epilepsy. Wimmer et al. used a so-called knock-in mouse model carrying a human mutation (p.Arg43Gln 5 R43Q) in the gene GABRG2 encoding the GABAA receptor g2 subunit. This mutation had been identified previously in a large Australian pedigree with variable phenotypes encompassing febrile seizures, childhood absence epilepsy, and other phenotypes. Functional studies of this mutation in oocytes, neurons, and in the aforementioned mouse model revealed a loss of function of the receptor, reduced surface expression, and reduced cortical inhibition in the primary somatosensory cortex (S1, also called barrel cortex). The mouse model nicely recapitulates the human phenotype with reduced thresholds for seizures induced by high temperature, mimicking febrile seizures, and spontaneous brief absence-like seizures with spike-and-wave discharges on the EEG. Here, Wimmer et al. studied the number of inhibitory and excitatory neurons in histologic examinations of the barrel cortex, where they previously found functional alterations; this was identified as the seizure initiation site in a different rodent model for absence epilepsy. Wimmer et al. chose a genetic background, in which they did not observe absence seizures and no spike-andwave discharges on EEG, to be sure not to detect morphologic changes secondary to absence seizures. Nevertheless, these mice showed a reduced threshold for thermally induced seizures, indicating a pathologic and hyperexcitable phenotype. Furthermore, they chose a young adult age (postnatal days 35–36 [P35–P36]) to detect all changes that may have happened during brain development. Surprisingly, the authors found an increase in the number of inhibitory GABAergic neurons and a reduced ratio of excitatory (counted as nonGABAergic) to inhibitory neurons, while the overall number of neurons was unchanged. They also detected layerand subtype-specific changes of inhibitory neurons in the S1 region. This finding is new and interesting, and it raises the question of whether morphologic changes could contribute to the development of epilepsy. However, its interpretation and the underlying mechanism behind this observation remain enigmatic. This also applies to other morphologic and functional imaging changes in genetic epilepsies reported so far. First of all, one would expect that an increase in GABAergic neurons should decrease neuronal excitability and seizure thresholds. However, it may not be as simple as that, since neuronal networks might be different with newly formed neurons and such changes may also have proconvulsive effects. Conversely, a simple explanation for the increase in the number of inhibitory neurons could be that it results from a compensation of the deficit of GABAergic inhibition in brain development. In this case, the compensation would not be sufficient to compensate for the inhibitory dysfunction. To answer such questions, many more studies at different time points may be helpful, but are tedious to perform. A second important question is whether the observed changes can be related to humans carrying the same or similar mutations, or to those individuals with IGE/GGE in whom morphologic changes have been described. For understandable reasons—i.e., to avoid a secondary effect of seizures—the authors used absence-naive animals. However, these mice may be different from those developing absences on a distinct genetic background. A solution, examining many different conditions with the same methods, would be necessary to answer such questions; however, to do this might necessitate automated analysis of morphologic investigations. The most robust morphologic change