Receptor trafficking hypothesis revisited: Plasticity of AMPA receptors during established status epilepticus

Authors


Address correspondence to Jaideep Kapur, Department of Neurology & Neuroscience, University of Virginia, Health Sciences Center, PO Box 800394, Charlottesville, Virginia, U.S.A. E-Mail: jk8t@virginia.edu

Summary

Status epilepticus (SE) is associated with a dynamic plasticity of postsynaptic neurotransmitter receptors. The plasticity of AMPA receptor (AMPAR)–mediated glutamatergic transmission during established SE (ESE), after development of benzodiazepine resistance, was evaluated. There was increased frequency and inward rectification of AMPAR-mediated excitatory postsynaptic currents at Schaffer collateral - CA1 pyramidal neuron synapses during ESE. Surface expression of the GluA1 subunit increased, and this was a consequence of N-methyl-d-aspartate receptor activation. Further, diminishing glutamate release by activation of somatostatin receptors prevented SE. These studies suggest that AMPAR-mediated glutamatergic transmission is strengthened during ESE.

Status epilepticus (SE), characterized by continual, self-sustained seizures, is a dynamic and rapidly evolving neurologic condition. As SE progresses, electrographic seizures become continuous, grade V behavioral seizures are observed in rats, and benzodiazepines (BDZs) fail to terminate seizures (e.g., Walton & Treiman, 1988; Kapur & Macdonald, 1997). This is an animal model of established SE (ESE). Understanding synaptic plasticity during ESE will help discover newer targets to treat BDZ-refractory SE. We investigated AMPA receptor (AMPAR)–mediated neurotransmission during ESE. We previously found that the expression of the GluA2 subunit of AMPARs in hippocampal principal neurons is dynamically reduced during ESE, leading to the expression of calcium-permeable AMPARs (Rajasekaran et al., 2012). Extending these findings, herein we hypothesize that AMPAR-mediated excitatory conductance is progressively enhanced during ESE. To test this hypothesis, SE was induced in lithium-pretreated adult male rats using pilocarpine, and the animals were studied either 10 min (early ESE) or 60 min (late ESE) after the onset of the first grade V behavioral seizure using a combination of electrophysiological and biochemical studies as described previously (Rajasekaran et al., 2012; Kozhemyakin et al., 2013).

AMPAR-mediated excitatory postsynaptic currents (EPSCs) were recorded from CA1 pyramidal neurons (CA1-PNs) and dentate granule cells (DGCs) by voltage-clamp technique. Analysis of recordings obtained from CA1-PNs revealed that the frequency of action-potential independent EPSCs (m-EPSCs) increased with increasing seizure duration (p < 0.0001, one-way analysis of variance [ANOVA]). The mean m-EPSC frequency in CA1-PNs of the control group was 0.39 ± 0.05 Hz (n = 14 cells/6 animals), whereas in the late ESE group, it was 1.59 ± 0.4 Hz (n = 12 cells/7 animals, p < 0.05, Tukey's test). In contrast, the mean m-EPSC frequency in CA1-PNs of the early ESE group was similar to controls (0.22 ± 0.04 Hz, n = 8 cells/6 animals). The amplitude of m-EPSCs in CA1 PNs from both early ESE (11.74 ± 0.8 pA) and late ESE (11.97 ± 0.8 pA) groups were similar to that in the control group (11.40 ± 0.6 pA); however, the net charge transfer of AMPAR-mediated m-EPSCs was significantly greater in the late ESE group (40.25 ± 6.1 pC vs. 138.9 ± 29.9 pC; p < 0.05, Tukey's test). Furthermore, when Schaffer collaterals were electrically stimulated to obtain a current-voltage (I-V) relationship, AMPAR-mediated evoked EPSCs (e-EPSCs) recorded from CA1-PN were found to be inwardly rectifying and philanthotoxin-sensitive during early ESE (Rectification Index, RI = 0.48 ± [SEM] 0.08 , n = 11 cells/5 animals) and late ESE (RI = 0.34 ± 0.06, n = 7 cells/6 animals). In contrast, the I-V relationship of e-EPSCs in control CA1-PNs did not show inward rectification or philanthotoxin sensitivity (RI = 0.87 ± 0.04, n = 7 cells/4 animals).

In contrast to CA1-PNs, recordings obtained from DGCs revealed no change in the frequency of m-EPSCs during early ESE (0.54 ± 0.5 Hz, n = 8 cells/5 animals) or late ESE (0.34 ± 0.6 Hz, n = 6 cells/4 animals) compared to that of control DGCs (0.59 ± 0.1 Hz, n = 6 cells/3 animals; p > 0.05, ANOVA). Likewise, the amplitudes of m-EPSCs in DGCs obtained from early ESE (11.5 ± 0.5 pA) and late ESE (11.31 ± 0.5 pA) animals were similar to that of control DGCs (12.6 ± 0.6 pA, p > 0.05, ANOVA). When e-EPSCs were obtained by stimulation of perforant path, the I-V relationship revealed inwardly rectifying, philanthotoxin-sensitive currents only during early ESE (0.56 ± 0.09, n = 12 cells/7 animals) but not late ESE (0.88 ± 0.09, n = 11 cells/7 animals). Control DGCs exhibited a linear I-V relationship (RI = 0.85 ± 10, n = 9 cells/5 animals). These studies showed that AMPAR-mediated synaptic transmission on CA1-PNs is strengthened from early ESE to late ESE.

We then determined changes in the surface expression of the GluA2 and GluA1 subunits of the AMPAR at early ESE and late ESE using a biotinylation assay to tag surface-expressed proteins. Compared to controls, the cell surface expression of the GluA2 subunit was reduced in hippocampi of early ESE (54 ± 15%, n = 8, p < 0.05, t-test) and late ESE (53 ± 6%, n = 6, p < 0.05, t-test) animals. In contrast to GluA2 subunit, the cell surface expression of the GluA1 subunit was increased in hippocampi of both early ESE (145 ± 6%, n = 6, p < 0.05, t-test) and late ESE (125 ± 10%, n = 3, p < 0.05, t-test) animals. The increased surface expression of the GluA1 subunit during SE was further confirmed using a BS3 cross-linking assay. In the hippocampi of late ESE animals, the intracellular fraction of the GluA1 subunit was 78 ± 5% of that in controls (n = 6 animals, p < 0.05, paired t-test). Because electrophysiological studies revealed distinct differences in the properties of AMPAR-mediated EPSCs between CA1-PN and DGCs during late ESE, a BS3 assay was performed on microdissected samples of the CA1 and DGCs. The intracellular fraction of GluA1 subunit was unchanged in the DGC subfield (104 ± 9%, n = 5, p > 0.05, t-test), whereas it was significantly reduced in the CA1 subfield (50 ± 10%, n = 7, p < 0.05, t-test). Biotinylation assay of the CA1 subfield also confirmed the increased surface expression of GluA1 subunit during late ESE (140 ± 12%, n = 4, p < 0.05, t-test). These studies indicate that strengthening of AMPAR-mediated neurotransmission on CA1-PN during ESE is associated with an increased surface expression of the GluA1 subunit.

An increase in cell surface expression of the GluA1 subunit during ESE indicated potential alterations in the trafficking of AMPARs. Activation of N-methyl-d-aspartate (NMDA) receptors (NMDARs) can also modulate surface expression of AMPARs; NMDARs are activated during SE. In preliminary studies, treatment of hippocampal slice cultures with NMDA (10 mm) and high extracellular potassium (10 mm) increased the surface expression of GluA1 subunit (193 ± 35%, n = 2). In addition, treatment of animals with NMDAR open-channel blocker, MK-801 (2 mg/kg) after 10 min of continuous electrographic seizures prevented increase in GluA1 surface expression in the CA1 region (97 ± 10%, n = 5, p > 0.05, t-test).

Posttranslational modification such as phosphorylation can also influence trafficking and conductance of AMPARs (Lu & Roche, 2012). In ongoing studies, we tested whether phosphorylation of the AMPARs at the Ser831 and Ser845 residue on the C-tail of the GluA1 subunit was increased during late ESE. Of interest, there were no differences in phosphorylation of either Ser831 (118 ± 15%, n = 8, p > 0.05, t-test) or Ser845 (117 ± 30%, n = 8, p > 0.05, t-test) residues in the whole hippocampi of animals during late ESE.

Because our studies demonstrate that glutamatergic transmission is enhanced during ESE, we tested whether diminishing glutamate release during SE can terminate seizures. Patch-clamp recordings obtained from CA1-PNs revealed that the neuropeptide, somatostatin (SST) reduced action potential-dependent EPSCs (s-EPSCs) and Schaffer collateral stimulated paired-pulse facilitation. These effects were inhibited by the SST type 2 receptor (SST2R) antagonist, cyanamid-154806 and were mimicked by the SST2R agonists, octreotide and lanreotide, suggesting that SST actions were mediated by SST2Rs. Video–electroencephalography (EEG) studies revealed that intraventricular administration of SST, within a range of doses, either prevented or attenuated pilocarpine-induced SE or delayed the median time to the first grade V seizure by 11 min. Similarly, octreotide or lanreotide prevented or attenuated SE in >65% of animals. Compared to the pilocarpine model, octreotide was highly potent in preventing or attenuating continuous hippocampal stimulation-induced SE in all animals within 60 min of SE onset (Kozhemyakin et al., 2013).

Together, these studies demonstrate that AMPAR-mediated synaptic transmission is enhanced during ESE and that reducing glutamate release is a potential therapeutic strategy to treat ESE.

Disclosure

The authors have no conflict of interest to declare in relation to this article.

The authors confirm that they have read the Journal's position on issues involved in ethical publication and affirms that this report is consistent with those guidelines.

Ancillary