Abstracts

Rationale: Hippocampal GABAergic inhibitory interneurons play various roles in the modulation of physiological network activity and in the modulation of epileptiform activity. Among the several GABAergic interneuron types, basket cells (BCs), which are classified based on their expression of either cholecystokinin (CCK) or parvalbumin (PV), innervate the somata and control the spiking output of pyramidal cells. Much more is known about the function of PV BCs than CCK BCs in vivo because of the existence of a PV-Cre mouse line. Until recently, similarly specific mouse line for CCK interneurons was not available. We show that an Sncg-Flp mouse line selectively labels CCK interneurons in control and epileptic mice, using immunostaining and in vivo two-photon calcium imaging.

Methods: Epilepsy was induced by intrahippocampal injection of kainate (KA; 70nl, 20mM in saline, -1.8 mm AP, 1.6 mm ML, -1.9mm DV from Bregma) or vehicle in Sncg-Flp mice of both sexes (3-5 months of age). A bipolar tungsten wire electrode was then implanted in the same site. 24hr video and EEG recordings were used to record seizures, and mice with 2 or more spontaneous behavioral seizures were considered epileptic. Flp-expressing cells were labeled using reporter viruses injected in the dorsal CA1 (bilaterally or contralateral to the KA side). For in vivo calcium imaging, a glass bottom cannula was implanted above the CA1 contralateral to the KA injection site, and two-photon calcium imaging was carried out in head-fixed mice freely running or resting on a linear treadmill. Finally, brains were perfused, sectioned, and immunostained against pro-CCK, and confocal z-stacks were acquired to quantify colocalization.

Results: Seizure monitoring in the chronic phase demonstrated that epilepsy could be reliably induced in Sncg-Flp mice, with the hallmarks of temporal lobe epilepsy. Virally labeled CA1 cells were located both ipsi- and contralaterally from the KA side, although ipsilateral cell numbers were reduced. Labeled cells were morphologically consistent with BC labeling, with axons forming a dense plexus in the pyramidal layer. Somata displayed a high degree of colocalization with pro-CCK. The in vivo activity pattern of the surviving contralateral cells was consistent with the characteristic response pattern of CCK BCs reported in control mice, as cells were suppressed during running and activated when stopped. Analysis of CCK BC activity during high frequency oscillations suggest that CCK interneurons are not recruited in these pathological network events, similarly to the suppression of CCK BCs during physiological sharp wave-ripple oscillations in controls.

Conclusions: Our results suggest that the Sncg-Flp transgenic mouse line allows specifically targeting CCK BCs in a chronic epilepsy model, enabling recording their activity in vivo, and, in future studies, selectively modulating CCK BC activity in epileptic mice. Furthermore, anatomical and functional evidence indicate that outside the locus of the epileptogenic insult, CCK BC form and function is largely preserved in chronic epilepsy. The role of this surviving CCK BC population in controlling seizure activity remains to be determined.

Funding: Please list any funding that was received in support of this abstract.: NIH NS99457.