Abstracts

Rationale: The GABAA γ2R43Q mutation confers Generalized Epilepsy with Febrile Seizures Plus (GEFS+) in humans (Wallace et al, 2001) and absence-like seizures in knock-in (RQ) mice (Tan et al, 2007). Humans harboring the γ2R43Q mutation show evidence of a hyperexcitable cortex compared to unaffected family members (Fedi et al., 2008), a condition hypothesized to contribute to the spike-and-wave discharges (SWDs) seen with absence seizures (Marini et al., 2003). The traditional first-line clinical anti-absence medication, ethosuxamide, works by decreasing T-type Ca2+ conductance in neurons (Hugeunard, 1999). This T-type Ca2+ conductance promotes ‘burst-firing' in thalamic relay neurons; a behavior inversely mediated by membrane potential and believed to be critical for SWD etiology (Cope et al., 2005, 2009). Interestingly, up to 40% of humans that suffer absence seizures are non-responsive to ethosuxamide (Tringham et al., 2012). Previous work in our lab shows RQ mice lack inhibitory tonic currents in both cortex and thalamus and, thus, may display altered neuronal firing and bursting behaviors. Methods: Thalamocortical slices (400 μm) (Krook-Magnuson et al., 2008) were cut from wild-type (RR) and RQ mouse brains and placed on an interface chamber perfused with low-Mg2+ artificial cerebral spinal fluid (aCSF). Multichannel extracellular recordings were made with two NeuroNexus 16-channel recording arrays placed in thalamus and cortex. Neurons were isolated using a principal components-based clustering and spike sorting algorithm. Home-written matlab code analyzed individual neuron spike-times for T-type Ca2+ ‘bursts', which included a 100 msec hyperpolarizing silence of activity followed by inter-spike-intervals of < 8 msec (Weyand et al., 2001). Each cell's tendency to fire in bursts was defined by a t-burst fraction: # of burst events divided by total # of events (bursts and lone spikes). T-burst fraction, spikes-per-burst, burst duration, and over-all firing rates were calculated and compared for each area with a Kruskal-Wallis examination of medians. Results: RQ cortical firing rates were increased compared to RR (Hz: median of means, IQR; RR: 0.04, 0.08, n=76; RQ: 0.06, 0.12, n=81; p<0.05), while RQ thalamic firing rates were decreased (RR: 0.12, 0.51, n=112; RQ: 0.04, 0.15, n=75; p<0.05). RQ mice thalamus showed a decreased t-burst fraction compared to RR (median, IQR; RR: 0.20, 0.25; RQ: 0.05, 0.15; p<0.05), a decreased number of spikes-per-burst (median spikes: RR: 3, 3; RQ: 2, 0; p<0.05), but no change in burst duration time (median milliseconds: RR: 6.2, 4.1; RQ: 6.1, 5.4). Similarly, RQ cortex also displayed a decreased t-burst fraction (RR: 0.14, 0.23; RQ: 0.05, 0.14; p<0.05), decreased number of spikes-per-burst (RR: 3, 3; RQ: 2, 0; p<0.05), and no change in burst duration time (RR: 5.4, 4.1; RQ: 3.9, 6.0). Conclusions: These results show that the γ2R43Q mutation increases firing rates in cortex and decreases T-type Ca2+ bursting in thalamus, findings consistent with a loss of inhibitory tonic currents in these neurons.