Abstracts

Rationale: Glutamatergic mossy cells (MCs) of the dentate gyrus (DG) are potentially critical to the DG because they can regulate GC activity through direct innervation of GCs or indirectly through GABAergic neurons. However, the net effect of MCs on DG and GC function, and the implications for epilepsy, remains a topic of debate. Here, we test the hypothesis that MC excitation of GCs is normally weak, but under special conditions found in epileptogenesis, MC excitation of GCs strengthens, 'switching' the role of MCs from primarily inhibitory and antiepileptogenic to the opposite. Methods: Dopamine receptor D2 (DrD2)-Cre+/- and Cre-/- mice were used to target MCs. Mice underwent stereotaxic surgery to inject virus encoding excitatory (e) or inhibitory (i) Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) into the 2 DG sites, bilaterally. Subdural electrodes were placed over the frontal cortex, hippocampus, and occipital cortex and 2-4wks later mice video EEG was recorded. DREADDs were activated with clozapine-N-oxide (CNO, 10mg/kg, s.c.) and 30 min later pilocarpine (250mg/kg, s.c.) was injected. Mice were recorded during status epilepticus (SE) and either perfused to analyze damage with Fluorojade 72hrs later or kept for video-EEG 1 month later. In other mice injected with virus encoding channelrhodopsin (ChR2), optogenetic activation of MC axons in the IML was used to study effects on GCs. Standard artificial cerebrospinal fluid (ACSF) was used and then excitability was increased (either individually or together: 0 mMMg2+, 5mM K+, the GABAA antagonist 10µM bicuculline, or GABAB antagonist,10µM CGP52432). Results: In vivo results revealed a bidirectional effect of MCs on SE. Activation of eDREADD resulted in rapid seizure onset (16.4±0.8min) compared to Cre-/- (35.8±2.7min) and iDREADD (44.2±3.38 min; p<0.02). Hippocampal power during SE was suppressed in iDREADD mice compared to eDREADD and Cre-/- mice. Inhibiting MCs during SE led to fewer seizures 24-48hrs later and less neuronal injury in hilus and CA3. Two wks video-EEG of chronic seizures showed reduced seizure number, frequency, and duration in iDREADD mice (vs. Cre-/-). Slice studies showed that ChR2 activation of MC axons in standard ACSF led to weak EPSPs in GCs but increasing excitability led to burst discharges. Remarkably, burst discharges propagated to CA3, followed by CA1 and subiculum. Conclusions: The results show a surprising bidirectional effect of MCs in a rodent model of epilepsy. Specifically, increasing MC activity through eDREADDs or optogenetic stimulation of MC axons facilitated seizure activity, whereas inhibiting MCs with iDREADDs attenuated SE and chronic epilepsy. These data suggest that under many conditions MCs may inhibit GCs via activation of GABAergic neurons, as previous studies suggest, but under some conditions the ordinarily weak excitatory effects of MCs on GCs can be greatly enhanced and have surprising consequences for seizures, neuronal damage, and chronic epilepsy. Funding: Natural Sciences and Engineering Research Council of Canada PDF to JBNew York State Office of Mental Health and NIH R01 MH-109305, 106983 to HS