Interaction of anticonvulsant drugs with presynaptic NMDA receptors in the rat entorhinal cortex

Jian Yang, Caroline Wetterstrand, Roland S.G. Jones
University of Bath, Bath, United Kingdom

We have previously shown that a number anticonvulsant drugs reduce the spontaneous release of glutamate at excitatory synapses in the entorhinal cortex (EC) by actions unrelated to blockade of voltage gated sodium or calcium channels (e.g. Cunningham et al., 2000; 2004). We have also demonstrated that glutamate release is tonically facilitated at EC synapses by presynaptic NMDA autoreceptors (Woodhall, et al., 2001). In the present experiments we determined whether phenytoin (PHT), gabapentin (GBP) or felbamate (FBM) might reduce glutamate release via blockade of the NMDA autoreceptor.

Spontaneous glutamate release was monitored via whole cell patch clamp recordings of spontaneous excitatory postsynaptic currents (sEPSCs) from neurones in layer V of the EC in brain slices prepared from 4-6 week old male Wistar rats. Postsynaptic NMDA receptors were blocked in the recorded neurones by inclusion of the channel blocker, MK801, in the patch pipette solution, allowing observation of drug effects at presynaptic NMDA receptors.

In 3 neurones, bath perfusion with the competitive NMDA antagonist, 2-AP5 (50 μM), decreased the frequency of sEPSCs from 6.5±2.9 Hz to 4.2±1.7 Hz. (mean±sem; P<0.01 paired t-test), reflecting blockade of the tonic facilitatory NMDA autoreceptor (Woodhall, et al., 2001). Subsequent addition of PHT (50 μM in the presence of 2-AP5) caused a further decrease in frequency to 3.4±1.1 Hz (P<0.05). In 4 further neurones the order of PHT and 2-AP5 applications was reversed. sEPSC frequencies were 7.8±2.5 Hz in control, falling to 4.6±1.1 Hz in PHT (P<0.01) and 2.4±0.8 Hz with addition of 2-AP5 (P<0.01). The lack of occlusion between the drugs suggests that PHT is not acting at presynaptic NMDA receptors. Similar results were obtained with GBP. In 3 neurones sEPSC frequency was reduced from 6.8±2 to 3.1±1.4 Hz by 2-AP5 and to 1.6±0.4 Hz with addition of GBP. In experiments where GBP preceded 2-AP5 (n=4) the respective frequencies were 4.1±1.1, 1.9±0.5 and 1.3±0.3 Hz. In contrast, the effects of FBM and 2-AP5 were occlusive. When 2-AP5 was applied first the frequency fell from 2.1±0.6 to 1.2±0.3 Hz (n=3; P<0.05), with no further reduction when FBM was added (1.3±0.3 Hz). Likewise when the drug order was reversed (n=4) the frequency decreased from 5.1±1.1 Hz to 2.5±0.9 Hz with FBP (P<0.01) but in 2-AP5 it did not change (2.6±0.9 Hz). sEPSC amplitudes were unaltered in any of the studies. Thus, the results suggest that FBM but not PHT or GBP can reduce glutamate release by blocking facilitatory NMDA autoreceptors. It is pertinent that we have shown that there is an increase in NMDA autoreceptor function at EC synapses in chronically epileptic rats (Yang et al., 2006), perhaps indicating a specific target for the anticonvulsant effect of FBM.

Cunningham, M.O. et al., (2000) Neuroscience. 95, 343-351
Cunningham M.O. et al., (2004) Eur J. Neuroscience. 20, 1566-1576
Woodhall G.L. et al., (2001) J. Neurophysiol. 86, 1644-1651
Yang et al., (2006) J. Neurosci. 26, 406-410

We thank the Wellcome Trust, Epilepsy Research UK for financial support.