Waking up the dormant dentate gyrus.

Commentary Identification and characterization of the vast variety of hippo-campal principal neurons and interneurons are necessary to understand the role that each cell type plays in the complex network alterations leading to epilepsy (1). Over several decades, research has focused on which cells are necessary, sufficient, or extraneous for hippocampal dysfunction often observed in animal models of epilepsy. In particular, temporal lobe epilepsy (TLE) is characterized by the death of many cells of the dentate hilus, and investigators have demonstrated that dentate granule cells (DGC) could be hyperexcitable (2) or hyperinhibited (3). In human TLE, it is also uncertain whether DGC are hyper-excitable (4, 5). A complete understanding of whether—and how—specific cell loss is related to epilepsy remains elusive; in particular, how the dentate gyrus (DG) contributes to epilepsy is uncertain since it is not clear that seizures begin there. A greater understanding of the role of the DG in epilepsy has been hampered by the inability to selectively delete or ablate specific cell types for selective analysis of the remaining circuitry. In particular, the role of mossy cells has been controversial (6). Mossy cells are glutamatergic hilar neurons that can either excite or inhibit DGC. An excitatory effect occurs via direct activation of DGC by mossy cells (feedback excitation); an inhibitory effect of mossy cells on DGC occurs when mossy cells excite inhibitory hilar interneurons, which in turn inhibit DGC by a " feedforward " mechanism. Owing to their location and physiological functions, hilar mossy cells have been predicted to play a crucial role in the regulation of hippocampal excitability after excitotoxic insults, such as status epilepticus and traumatic brain injury. Therefore, the consequences of mossy cell loss are key to understanding the pathophysiology of TLE. Possible explanations include the " dormant basket cell hypothesis " (7), which holds that mossy cell death eliminates the feedforward inhibition by disinhibiting granule cells and allows them to fire repetitively (which they do not ordinarily do); as a consequence, there is increased circuit excitability and seizures. Conversely, the " irritable mossy cell hypothesis " (6) posits that mossy cells surviving an insult (such as status epilepticus) provide excessive excitatory innervation to granule cells, leading to their hyperexcitability. Finally, the role of mossy fiber sprouting in epileptogenesis remains uncertain; it is possible that in addition to autoinnervation of granule cells, mossy fibers might also sprout onto mossy cells and interneu-rons, altering the excitability …