Abstracts

Rationale: Hilar mossy cells of the dentate gyrus innervate dentate granule cells and local interneurons to modulate dentate function, but whether these cells produce a net excitatory or inhibitory effect is a topic of considerable debate. Determining the functions of mossy cells is especially important in the context of epilepsy since these cells are extremely vulnerable to insults such as seizures. Here, we test the hypothesis that mossy cells are normally inhibitory, but at the onset of status epilepticus (SE), mossy cells produce excitatory effects within the dentate gyrus. Methods: We used dopamine D2 receptor transgenic adult male mice that selectively express cre-recombinase in hilar mossy cells (DrD2-Cre). For in vivo studies, DrD2-Cre +/- mice (n=15) underwent stereotaxic surgery at 8 weeks of age to inject the inhibitory DREADD AAV2-hSyn-DIO-hM4di (iDREADD mice). Controls (n=15) were DrD2-Cre -/- mice injected with iDREADD or AAV2-DIO-YFP. To target the majority of mossy cells and their terminals, we injected virus (150nL) in the dorsal and ventral hippocampus, bilaterally. Subdural screws positioned over the left frontal, left and right dorsal, and right ventral hippocampus were used for video-EEG recordings. After >2 weeks, all mice were connected to a video-EEG system for continuous monitoring and pre-treated with 10mg/kg CNO (s.c.). Thirty minutes later, mice received 250 mg/kg pilocarpine (s.c.) to induce SE. Approximately half of the iDREADD and control mice (n=15) were euthanized 3 days later to evaluate neuronal injury with FluoroJade B. The other half of the mice (n=15) stopped video-EEG monitoring for >4 weeks and then were recorded continuously for 2 weeks to evaluate chronic spontaneous seizures (epilepsy). For in vitro studies, mice were injected with 150nL AAV-DIO-ChR2 and used for slice electrophysiology >2 weeks later. ChR2-expressing mossy cell terminals were activated with blue light under normal (3.5mM K⁺, 1.0mM Mg²⁺) or seizure-like (5mM K⁺, =0.5mM Mg²⁺, 10µM bicuculine) ACSF conditions, while simultaneously recording extracellular field potentials in the dentate gyrus or area CA3.  Results: iDREADD silencing of mossy cells prior to pilocarpine delayed the latency to the first convulsive seizure relative to controls. The latency to SE was not different, but the development of the most severe stage of SE was delayed in iDREADD mice. There was a reduction in EEG spectral power during the first 30 minutes of SE in iDREADD mice. iDREADD mice also exhibited differences from controls in the days and weeks after SE. For example, there were fewer convulsive seizures on the day after pilocarpine in iDREADD mice. iDREADD mice demonstrated reduced neuronal injury within the hippocampus when Flourojade B was used 3 days after SE. Moreover, there were fewer and less severe chronic seizures in iDREADD mice. Slice recordings showed that ChR2-activation of mossy cell axons had little excitatory effect normally but when seizure-like conditions were used, there was strong excitation of granule cells and this propagated to area CA3 pyramidal cells.  Conclusions: The results suggest that inhibition of mossy cells protects against SE, SE-induced hippocampal damage, and SE-induced chronic seizures. We suggest that mossy cells have potent excitatory drive on granule cells when there are conditions that either simulate a seizure or are during the onset of pilocarpine-induced SE. Under these conditions granule cells are depolarized, GABA release is transiently depleted This conditional, excitatory role of mossy cells is a contrast to normal physiological conditions, where mossy cells appear to mostly inhibit granule cells. Funding: National Institutes of Health (HS&HB) & NSERC (JB)