Abstracts

Rationale: Autosomal dominant mutations (S326fs328X, A322D) in the GABA_A receptor (GABA_AR) α1 subunit are associated with human absence epilepsy and juvenile myoclonic epilepsy, respectively, and both these mutations substantially reduce α1 subunit protein expression in vitro. Recently, we reported that heterozygous deletion of the α1 subunit (Het_α1KO) caused electrographic and behavioral absence seizures in mice. In addition, we reported that although Het_α1KO reduced total α1 subunit protein by approximately 50%, it also increased α3 subunit expression as well as the surface fraction of total α1 subunit and thereby altered the amplitude and kinetic time course of miniature inhibitory postsynaptic currents (mIPSCs). Here, we determined whether these changes in GABA_AR surface expression resulted from GABA_AR recruitment from the endoplasmic reticulum (ER), or from reduced endocytosis from the cell surface. Methods: We first determined if Het_α1KO altered the fraction of total α1 subunit localized in the ER. We prepared protein lysates from wild type and Het_α1KO cortices, and incubated them either with endoglycosdiase-H, an enzyme that selectively cleaves high mannose N-linked glycans associated with ER-resident proteins, PNGase F, an enzyme that cleaves all N-linked glycans, or no glycosidase (control), and analyzed the digestion products on Western blot. To elucidate the effect of Het_α1KO on GABA_AR endocytosis, we performed patch-clamp electrophysiology studies on cortical layer VI pyramidal neurons and recorded mIPSCs. After a five minute baseline period, we added Dynasore, an inhibitor of clathrin-mediated endocytosis, and determined whether endocytosis inhibition altered mIPSCs differently in wild type and Het_α1KO neurons. Results: Surprisingly, in contrast to recombinant GABA_AR overexpressed in heterologous cells, neither wild type nor Het_α1KO cortices possessed any detectable ER-resident α1 subunit, a result demonstrating that ,Het_α1KO does not recruit ER receptors to increase the fraction of total α1 subunit on the cell surface. Dynasore caused a time-dependent increase in mIPSC amplitudes in wild type neurons that reached 28% at 30 minutes. In contrast, Dynasore had no effect on the mIPSCs from Het_α1KO neurons, a result consistent with tonically reduced GABA_AR endocytosis. Conclusions: Het_α1KO does not recruit ER-resident receptors to increase the surface fraction of total α1 subunit. However, Het_α1KO does alter surface GABA_AR expression, at least in part, by reducing endocytosis. A tonic reduction of the rate of GABA_AR endocytosis would prevent Het_α1KO neurons from dynamically making further reductions in endocytosis to increase synaptic inhibition on-demand. A deficit in this dynamic mechanism to fine-tune synaptic inhibition could contribute to the development of paroxysmal hyperexcitability and seizures.