Abstracts

Rationale: Epilepsy patients must cope with a host of cognitive and emotional comorbidities, which many patients find more debilitating than the seizures. However, the neural mechanisms mediating these cognitive impairments in temporal lobe epilepsy (TLE) are not well defined. Pattern separation, one of the fundamental processes of cognition, is mediated by remarkably sparse activation in hippocampal dentate granule cells (DGCs). Sparse DGC activity has a secondary effect: it limits cortical input to the hippocampus and acts like a "gate," the failure of which may contribute to the excessive cortical-hippocampal activity underlying TLE seizures. Our laboratory previously found a more than 10-fold increase in DGC activity following epilepsy onset in a pilocarpine TLE mouse model. In this study, we used Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) both to hyperactivate DGCs in otherwise normal mice and to reduce DGC activity in epileptic mice to isolate the potential cognitive consequences of DGC activity on a hippocampal memory task. Methods: The spatial object recognition (SOR) task, which assesses pattern separation, introduced mice to 3 objects; 24 hours later, 1 object was moved, and the amount of time the mice explored the displaced object (DO) reflected if the mice recognized a change in their spatial environment (Wilcoxon's signed rank test). DGC activity was manipulated with virally expressed DREADDs (AAV5.HA.CamKIIa.hM3d/hM4d.IRES.mCitrine) in the dorsal DG. Epileptic mice expressing the inhibitory DREADD (hM4d) received 1.5 mg/kg CNO i.p. 1 hr prior to training and testing, while healthy mice expressing the excitatory DREADD (hM3d) received saline or 0.3 mg/kg CNO i.p. 1 hr prior to the test trial. Voltage sensitive dye (VSD) recordings in coronal DREADD-expressing slices examined DGC responses to stimulation with and without CNO. Results: Control mice exhibited robust pattern separation, spending significantly more time exploring the DO compared to the non-displaced objects (NDOs; n=8, p=0.016). However, 2-3 month post-status mice failed to discriminate (n=5, p=0.844). DREADDs reversed SOR performance in these groups. CNO-activated hM4d restored pattern separation in epileptic mice (n=2, 54% increase in DO preference). CNO-treated healthy mice expressing hM3d failed to discriminate (n=7, p=0.578), while their saline-treated controls were successful (n=7, p=0.015). To confirm that DREADDs were appropriately modulating excitability, hM3D-expressing mice were a FosTRAP transgenic line in which fos drives tdTomato expression. FosTRAPed cells provided a "snapshot" of the number of DGCs recruited in vivo and confirmed that CNO did not induce spontaneous seizure-like activity. VSD confirmed CNO-dependent changes in DGC recruitment. Conclusions: These data collectively suggest that DGC hyperactivation, as seen in epilepsy, is sufficient to compromise behavioral performance, and that reducing DGC hyperactivity in epileptic mice is sufficient to restore performance. Thus, DGC hyperactivity is a critical factor in TLE cognitive comorbidities. Funding: NIH grants to DAC (R01 NS082046) and JBK (T32 MH17168)