(13) Levetiracetam preferentially acts at neuronal synapses that only express SV2A

Rationale: Levetiracetam (LEV) is a third-generation frontline antiseizure drug (ASD) that lacks many of the adverse side effects of classic ASDs. LEV binds the synaptic vesicle protein, SV2A, and has been shown to decrease both excitatory and inhibitory neurotransmission in hippocampal slices (Meehan et al, Epilepsia 2012). The molecular basis of the anticonvulsant action of LEV is unknown, however, and direct effects of LEV on SV2A function remain to be fully established. SV2A is the most widely expressed paralog of a three-gene family (SV2A, B, C) that is variably co-expressed throughout the CNS. However, only SV2A is expressed in the majority of GABAergic neurons and dentate granule cells of the hippocampus, two classes of neurons implicated in the excitability in epilepsy. The major objective of this study was to thus define the extent to which the presence of non-LEV binding SV2 paralogs, i.e. ,SV2B, influences LEV action in dissociated neurons and in intact mice.

Methods: Excitatory principal neurons from the CA region of hippocampus were cultured from SV2B-/- (BKO) mice to define the action of LEV on synaptic depression, an indicator of high neurotransmitter release probability. Single neurons were grown on microislands of astrocytes allowing them to form autaptic synapses. Cultures were treated +/- 300uM LEV for 3-7 hours; a concentration and incubation time that can alter synaptic transmission in hippocampal slices (Meehan et al, J Neurophysiol 2011). Excitatory postsynaptic currents (EPSCs) in response to 2 sec, 20 Hz stimulus trains were then recorded. We next quantified the dose-related efficacy of LEV against evoked 6 Hz 32 and 44 mA focal seizures in adult BKO and wild-type (WT) male mice (n=6-10/group), and against maximal electroshock seizures (MES) at a high dose (500 mg/kg, i.p.).

Results: LEV selectively reduced synaptic depression only in dissociated excitatory hippocampal neurons from BKO mice. It had no effect on WT hippocampal neurons. LEV also increased the paired pulse ratio in neurons from BKO mice and had no effect in cultured WT neurons. In vivo, LEV demonstrated dose-related anticonvulsant efficacy in both the 6 Hz 32 and 44 mA test in BKO mice; an effect that was not detected in WT mice at the doses tested (70 and 140 mg/kg, i.p.). In the 6 Hz 32 mA test, LEV (140 mg/kg, i.p.) protected 8/8 BKO mice from focal seizure, whereas only 2/8 WT mice were protected. In the 6 Hz 44 mA test, LEV (140 mg/kg, i.p.) protected 3/8 BKO mice from focal seizure, whereas 0/9 WT mice were protected. LEV (500 mg/kg) was without anticonvulsant effect in the MES test in both BKO and WT mice.

Conclusions:

Our present mechanistic and behavioral studies suggest that LEV acts preferentially at synapses that only express SV2A. The in vivo efficacy of LEV was most notable in BKO mice, and this efficacy was preserved in a dose-related manner against the drug-resistant 6 Hz 44 mA seizure. LEV may preferentially act at synapses critical to evoked focal, but not generalized, seizures. Future studies must define whether the selective action of LEV in discrete neuronal subpopulations is necessary and sufficient for LEV anticonvulsant efficacy.


Funding: Please list any funding that was received in support of this abstract.: This work was supported by the University of Washington Department of Pharmacy (MBH), the Institute for Translational Health Sciences (ITHS KL2 TR002317 to MBH), and the University of Washington School of Medicine Bridge Fund (SMB).