Abstracts

Rationale: The dentate gyrus is an integral part of the trisynaptic circuit involved in episodic memory formation by processing complex cortical inputs into sparse activity in granule cells (GCs), the canonical projection neuron. Signal sparsification is mediated by a precise spatial and temporal regulation of excitation by inhibition. Temporal lobe epilepsy is known to cause extensive reorganization of dentate circuits and contribute to memory impairments by mechanisms that are not fully understood. A unique subset of glutamatergic projection neurons termed the semilunar granule cell (SGC), with expansive molecular layer dendritic arbors, has been proposed to shape memory processing by supporting feedback inhibition (Larimer et al., 2010, Afrasiabi et al., 2022) and is preferentially recruited in memory representations (Erwin et al., 2020). However, whether SGC intrinsic physiology and synaptic inputs are altered in epilepsy, as shown in concussive brain injury (Gupta et al., 2012), is not known. This study was conducted to identify the changes in SGC intrinsic and synaptic physiology in epilepsy.

Methods: Adult male and female (4-7 weeks) C57BL/6 mice, pilocarpine-induced status epilepticus (SE) animals or saline-injected and naïve controls, were euthanized 1-month post-SE for slice physiology. Whole-cell current and voltage recordings from GCs and SGCs were obtained from horizontal hippocampal slices (350 µm). Recordings were obtained using K-gluconate internal to examine intrinsic physiology and Cs-methanesulfonate to examine excitatory and inhibitory postsynaptic currents (EPSCs and IPSCs). Recorded cells were filled with biocytin and stained for cell identification. Cells were reconstructed using Neurolucida 360 to analyze morphology and spine density. 

Results: SGCs show a higher density of dendritic spines and received more frequent sEPSCs compared to GCs (Frequency in Hz, GC: 0.8 ± 0.14, SGC: 2.3 ± 0.64, p< 0.05 in 10 GC and 8 SGC) in naïve mice. Frequency and amplitude of sEPSCs in both cell types increased following experimental epilepsy. In parallel, SGCs also received significantly larger sIPSCs after experimental SE (in pA, SE SGC: 72.48 ± 1.07, Sham SGC: 63.79 ± 1.03 Sham n=4, SE n=9 p< 0.05; Cohen’s d= 0.46) not observed in GCs falling in line with previously demonstrated differences in sIPSC frequency between GCs and SGCs in naïve mice. Interestingly, intrinsic physiology of SGCs including resting membrane potential and input resistance were not different between sham and mice after SE (n= 3-6 cells/group).