Abstracts

Rationale: The dentate gyrus (DG) is critically involved in cognitive processes such as pattern separation and is crucial to the formation of episodic memories. Dentate granule cells (DGCs) exhibit sparse activation of distinct subsets of populations during completion of cognitive tasks in vivo. The mechanisms mediating the sparse, but selective activation of DGCs are unknown. The DG also plays an important role in the regulation of pathological activation of the limbic system, functioning as a regulated gate, restricting relay of synchronous network activity associated with epilepsy. Epilepsy significantly erodes DG circuitry resulting in recurrent mossy fiber sprouting, changes in local inhibition, and aberrant neurogenesis. In this study, we examined how the DG is altered during epilepsy by stimulating DGC afferents across different frequency ranges encompassing those which activate the circuit in vivo during cognitive processing.Methods: We utilized fluorescent calcium indicators coupled with both two-photon and fast-confocal microscopy to conduct multicellular calcium imaging of DGCs responding to perforant path stimulation at 1, 5, 10, and 60 Hz frequencies in hippocampal entorhinal-cortex slices from control and pilocarpine model epileptic adult mice. Mice were 18 weeks old, 8 weeks post status epilepticus and were verified as chronically epileptic by behavioral monitoring.Results: Stimulation of DGC afferents at low frequencies elicited APs in a sparse population of DGCs (4.4% for 1 Hz and 6.2% for 5 Hz, n=543 cells), mimicking activation observed in vivo. Interestingly, stimulation at higher frequencies activated significantly more DGCs, (15.2% for 10 Hz and 28.9% for 60 Hz). Epileptic mice showed a robust increase in proportional DGC activation and a loss of frequency dependent recruitment [stimulus trains of 1, 5, 10 and 60 Hz all produced similar proportional activation of DGCs (43%, 55%, 50%, 53%, respectively, n = 477 cells)]. For all frequencies tested, DGCs activated reliably to repeated stimulation. Lastly, we explored how DGCs responded to stimulation at multiple sites and whether distinct stimuli activated unique ensembles of DGCs. We found that there was significant overlap in DGCs activated mutually to two distinct stimuli in both control (59.9%) and epileptic mice (55.0%), suggesting deterministic firing. Moreover, these recordings from epileptic mice, revealed a robust increase in proportional DGC activation compared to a much more restricted activation profile seen in control animals. Conclusions: DGC activation is normally sparse, reliable, frequency dependent, and deterministic. Many of these characteristic DGC network activation properties exhibit significant epilepsy-associated deterioration. These results provide insight in understanding basic hippocampal network functions and how they are altered during epilepsy, and may be helpful in developing new treatments and therapies for epilepsy and its co-morbidities.