Kainate receptor mediated facilitation of GABA release and slow wave oscillations in the rat entorhinal cortex in vitro.

Sophie Chamberlain, Roland Jones

University of Bath, Bath, UK

Rhythmical fluctuations in membrane potential of cortical neurones or slow wave oscillations (SWO) are a hallmark of deep sleep and thought to have a critical role learning-induced plasticity. SWO can be elicited in slices of rat entorhinal cortex (EC) by a moderate reduction in [Mg2+]o and one factor involved is increased activation of a GluR5-subunit containing kainate receptor (KAr), since SWO can be blocked by the selective antagonist, UBP 302 (Cunningham et al., 2005). We have shown previously that GluR5 KAr act as facilitatory autoreceptors at glutamate synapses in the EC. This function is enhanced during SWO generation (Chamberlain and Jones, 2007). Both excitatory and inhibitory transmission increase during SWO (Cunningham et al., 2005). In the present study we determined the contribution of KAr to spontaneous GABA release, and the effect of lowering [Mg2+]o in EC slices prepared from male Wistar rats (60-70g). Spontaneous inhibitory postsynaptic currents (sIPSCs) mediated by GABAA receptors were recorded using whole cell-patch clamp from layer III neurones in the presence of specific AMPA and NMDA receptor antagonists, and used as a reporter of presynaptic GABA release.

In ‘normal’Mg2+ (2 mM) the frequency of sIPSCs was 15.4±3.2 Hz (n=7). Application of UBP 302 (20 μM) reduced sIPSC frequency to 10.1±2.1% Hz (mean reduction -31.3 ± 5.8 %). Subsequent addition of the non-selective KAr antagonist, CNQX (10 μM), caused a further decrease in frequency to 7.1±0.9 Hz (-20.1±10.5 %). Blockade of action potential driven release by perfusion with TTX (1 μM) reduced sIPSC frequency by around 60-80% in most layer III neurones. The frequency of miniature IPSCs (mIPSCs) in 6 neurones recorded in these conditions was 5.0±2.1 Hz, and this was unaffected by UBP 302 (5.2±2.2 Hz; +7±4%). In a subset of three neurones, control mIPSC frequency was 3.1±1.3 Hz and this was unaffected by UBP 302 (3.0±1.1 Hz; +1.3±4.0%) but reduced by subsequent addition of CNQX (1.5±0.5 Hz; -44.7±11.3%). In 8 neurones, reduction of [Mg2+]o from 2 to 1.25 mM increased sIPSC frequency from 9.9±1.7 Hz to 15.3±1.8 Hz (+76±41 %). In 4 of these neurones, sIPSC frequency increased from 8.5±1.8 Hz to 19.8±1.6 Hz in low Mg2+, but addition of UBP 302 then lowered this to 13.4±1.5 Hz (-37±2.6%). Frequency was further reduced to 12.2±1.4 Hz (-10±0.8%) with the addition of CNQX, but remained elevated above control levels (+24.3±5.8%). sIPSC and mIPSC amplitudes were not significantly affected by either antagonist in any of these studies.

These data suggest that spontaneous GABA release onto layer II neurones is partly driven by excitation of interneurones via a glutamatergic drive mediated by GluR5-containing KAr. It is further facilitated by spillover of glutamate to a non-GluR5 containing receptor on the GABA terminals. GABA release is elevated under conditions that generate SWO, and this may be partly due to enhanced activation of the KAr on interneurones, subsequent to enhanced glutamate release (Chamberlain et al, 2008).

Chamberlain, S.E.L. et al., (2007) Proceedings of the British Pharmacological Society at http://www.pA2online.org/abstracts/Vol5Issue2abst004P.pdf

Cunningham, M.O. et al., (2006) P.N.A.S. 103, 5597-5601