No abstract

No abstract

We have previously shown that presynaptic N-methyl-D-aspartate receptors (NMDARs) can facilitate glutamate release onto principal neurons in the entorhinal cortex (EC). In the present study, we have investigated the subunit composition of these presynaptic NMDARs. We recorded miniature alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor-mediated excitatory postsynaptic currents (mEPSCs), from visually identified neurons in layers II and V of the EC in vitro. In both layers, bath application of the NR2A/B subunit-selective agonist, homoquinolinic acid (HQA), resulted in a marked facilitation of mEPSC frequency. Blockade of presynaptic Ca(2+) entry through either NMDARs or voltage-gated Ca(2+) channels with Co(2+) prevented the effects of HQA, confirming that Ca(2+) entry to the terminal was required for facilitation. When the NR2B-selective antagonist, ifenprodil, was applied prior to HQA, the increase in mEPSC frequency was greatly reduced. In addition, we found that an NMDAR antagonist blocked frequency-dependent facilitation of evoked release and reduced mEPSC frequency in layer V. Thus we have demonstrated that NMDA autoreceptors in layer V of the EC bear the NR2B subunit, and that NMDARs are also present at terminals onto superficial neurons.

We have shown previously that when postsynaptic NMDA receptors are blocked, the frequency, but not amplitude, of spontaneous EPSCs (sEPSCs) at synapses in the entorhinal cortex is reduced by NMDA receptor antagonists, demonstrating that glutamate release is tonically facilitated by presynaptic NMDA autoreceptors. In the present study, we recorded sEPSCs using whole-cell voltage clamp in neurons in layer V in slices of the rat entorhinal cortex. Using specific antagonists for NR2A [(R)-[(S)-1-(4-bromo-phenyl)-ethylamino]-(2,3-dioxo-1,2,3,4-tetrahydroquinoxalin-5-yl)-methyl]-phosphonic acid] and NR2B [(␣R,␤S)-␣-(4-hydroxyphenyl)-␤-methyl-4-(phenylmethyl)-1-piperidinepropanol hydrochloride (Ro 25-6981)] subunit-containing receptors, we confirmed that in slices from juvenile rats (4 -6 weeks of age), the autoreceptor is predominantly of the NR1-NR2B subtype. In older (4 -6 months of age) control animals, the effect of the NR2B antagonist was less marked, suggesting a decline in autoreceptor function with development. In slices from rats (aged 4 -6 months) exhibiting spontaneous recurrent seizures induced with a lithium-pilocarpine protocol, Ro 25-6981 again robustly reduced sEPSC frequency. The effect was equal to or greater than that seen in the juvenile slices and much more pronounced than that seen in the age-matched control animals. In all three groups, the NR2A antagonist was without effect on sEPSCs. These results suggest that there is a developmental decrease in NMDA autoreceptor function, which is reversed in a chronic epileptic condition. The enhanced autoreceptor function may contribute to seizure susceptibility and epileptogenesis in temporal lobe structures.

We have recently shown that the anticonvulsant drugs phenytoin, lamotrigine and sodium valproate all reduce the release of glutamate at synapses in the entorhinal cortex in vitro. In the present investigation we determined whether this property was shared by gabapentin and pregabalin, using whole-cell patch-clamp recordings of excitatory postsynaptic currents (EPSCs) in layer V neurons in slices of rat entorhinal cortex. Both drugs reduced the amplitude and increased the paired-pulse ratio of EPSCs evoked by electrical stimulation of afferent inputs, suggesting a presynaptic effect to reduce glutamate release. The frequency of spontaneous EPSCs (sEPSCs) was concurrently reduced by GBP, further supporting a presynaptic action. There was no significant change in amplitude although a slight reduction was seen, particularly with gabapentin, which may reflect a reduction in the number of larger amplitude sEPSCs. When activity-independent miniature EPSCs were recorded in the presence of tetrodotoxin, both drugs continued to reduce the frequency of events with no change in amplitude. The reduction in frequency induced by gabapentin or pregabalin was blocked by application of the l-amino acid transporter substrate l-isoleucine. The results show that gabapentin and pregabalin, like other anticonvulsants, reduce glutamate release at cortical synapses. It is possible that this reduction is a combination of two effects: a reduction of activity-dependent release possibly via interaction with P/Q-type voltage-gated Ca channels, and a second action, as yet unidentified, occurring downstream of Ca influx into the presynaptic terminals.