Temporal lobe epilepsy (TLE) is among the most common forms of epilepsy, accounting for approximately 30% of cases. It follows a characteristic course: an initial insult—status epilepticus, traumatic brain injury, encephalitis, or prolonged febrile seizures—triggers the transformation of a healthy brain into an epileptic one. This is followed by a so-called latent phase, free of clinical seizures, lasting from a few weeks to several years, before entry into the chronic phase marked by recurrent seizures. The molecular and cellular mechanisms underlying this transformation, collectively termed epileptogenesis, remain incompletely understood. This issue is all the more important given that approximately 30% of patients develop drug-resistant epilepsy.
To map these dynamics, the authors applied spatial transcriptomics, a sequencing technology that enables profiling of the entire transcript repertoire while preserving its localization on the histological section. The analysis focused on coronal brain sections from rats subjected to a classic model of status epilepticus induced by lithium-pilocarpine, compared with controls (n = 16), at various time points spanning the latent and chronic phases. Spatial clustering of the data allowed precise delineation of anatomically relevant regions and recovery of the expected canonical markers in controls, thereby validating the analytical approach.
Comparative analyses reveal substantial transcriptional alterations associated with status epilepticus, extending into the latent and chronic phases across all examined regions. The most striking finding is that microglial activation and reactive astrogliosis are not confined to the hippocampus: from the latent phase onward, they extend to white matter tracts and several thalamic nuclei. Cell-type deconvolution further highlights marked transitions between functional astrocyte subtypes within these reactive zones. These observations unveil the spatial heterogeneity of epileptogenic processes and underscore a previously underappreciated involvement of the thalamus and white matter beyond the classically recognized limbic structures.
The authors acknowledge several limitations. The lithium-pilocarpine model in young adult rats does not reproduce the full complexity of human epilepsy, and the administration of lithium or diazepam may itself alter excitatory/inhibitory homeostasis and transcriptional profiles, making it difficult to distinguish drug effects from changes intrinsic to epileptogenesis. Moreover, the analytical methods, originally designed for single-cell sequencing, do not fully exploit spatial information. The authors conclude that the demonstration of region-specific glial responses and astrocyte subtype transitions calls for therapeutic strategies targeted by region and by cell type.