|Home | About | Journals | Submit | Contact Us | Français|
Removing Entorhinal Cortex Input to the Dentate Gyrus Does not Impede Low Frequency Oscillations, an EEG-biomarker of Hippocampal Epileptogenesis.
Meyer M, Kienzler-Norwood F, Bauer S, Rosenow F, Norwood BA. Sci Rep 2016;6:25660. [PubMed]
Following prolonged perforant pathway stimulation (PPS) in rats, a seizure-free “latent period” is observed that lasts around 3 weeks. During this time, aberrant neuronal activity occurs, which has been hypothesized to contribute to the generation of an “epileptic” network. This study was designed to 1) examine the pathological network activity that occurs in the dentate gyrus during the latent period, and 2) determine whether suppressing this activity by removing the main input to the dentate gyrus could stop or prolong epileptogenesis. Immediately following PPS, continuous video-EEG monitoring was used to record spontaneous neuronal activity and detect seizures. During the latent period, low frequency oscillations (LFOs), occurring at a rate of approximately 1 Hz, were detected in the dentate gyrus of all rats that developed epilepsy. LFO incidence was apparently random, but often decreased in the hour preceding a spontaneous seizure. Bilateral transection of the perforant pathway did not impact the incidence of hippocampal LFOs, the latency to epilepsy, or hippocampal neuropathology. Our main findings are: 1) LFOs are a reliable biomarker of hippocampal epileptogenesis, and 2) removing entorhinal cortex input to the hippocampus neither reduces the occurrence of LFOs nor has a demonstrable antiepileptogenic effect.
In healthy tissue, dentate granule cells respond to entorhinal input with very sparse activation, limiting the flow of excitation through the hippocampal circuit. This ‘gating’ function of the dentate can be impaired in temporal lobe epilepsy, such that even normal input from the entorhinal cortex might produce greater levels of activity in the dentate. This greater dentate output would then feed through the circuit and be proictogenic. A recent study by Norwood and colleagues adds an interesting twist to the classic dentate gate hypothesis of temporal lobe epilepsy; their findings reported in Meyer et al. suggest that entorhinal input may not be necessary for ictogenesis.
Meyer et al. used the perforant path stimulation model to induce temporal lobe epilepsy in rats. In this model, after 2 priming days of 30-min stimulation, the perforant path is stimulated for 8 hours in awake animals at a subconvulsive level . This results in immediate hippocampal cell damage and loss, and the appearance of spontaneous dentate discharges. Spontaneous behavioral seizures arise in this model 2 to 3 weeks after the 8-hour performant path stimulation insult.
The spontaneous discharges in the dentate, which appear immediately after the 8 hours of perforant path stimulation are remarkably similar to potentials directly evoked by stimulation of the entorhinal input to the dentate. This similarity caused Norwood and colleagues to previously note “the first spontaneous potentials closely resembled what would be expected from the granule cell layer if the seizures originated in the entorhinal cortex” and, holding with the classic view of the dentate gate hypothesis, “the available evidence is consistent with spontaneous ‘hippocampal-onset’ granule cell seizure discharges that may have been preceded and driven by spontaneous and synchronized entorhinal cortex discharges” (1). Thus, a reasonable interpretation was that the spontaneous discharges in the dentate were driven by the entorhinal cortex. A similar scenario was suggested for the spontaneous behavioral seizures that later develop in the model: “We hypothesize that the seizures in this model begin outside the hippocampus… and that only when the originating seizure activity recruits epileptiform discharges from disinhibited dentate granule cells do clinical seizures occur” (1). Previous findings were thus consistent with the hypothesis that input from the entorhinal cortex ultimately drove seizure expression. And so, in Meyer et al., they did a simple test–immediately after perforant path stimulation, they physically cut off the entorhinal input to the dentate. The effect? Nothing.
Severing the perforant path had no effect on the appearance or frequency of the spontaneous discharges in the dentate. It also had no effect on the time to the first large behavioral seizure. Nor did it impact the severity of that seizure. Even if there were some residual entorhinal input, or some reinnervation over time, the complete lack of measurable change is striking, and requires a reconsideration of the classical formulation of the dentate gate hypothesis.
There are two obvious possibilities. One is that the input to the dentate that is not being properly ‘gated’, the input that causes the seizure, is not of entorhinal origin. While the entorhinal cortex is the major excitatory input to the dentate (2, 3), it is certainly not the only input. The dentate gyrus receives input from the presubiculum and parasubiculum; the postrhinal and perirhinal cortex; backprojections from CA1 and CA3; commissural connections; GABAergic & cholinergic, noradrenergic, dopaminergic, and serotonergic input from the septum, locus coeruleus, ventral tegmental area, and raphe; and GABAergic and glutamatergic input from the supramammillary nucleus (3, 4, 5). One or more of these other sources of input (specifically, those that would not be damaged by perforant path transection) to the dentate may drive seizures. In this scenario, the dentate gate hypothesis requires only overt acknowledgement of these other sources of input. However, this input would have to be sufficiently impactful that removal of entorhinal input would have no measurable effect on dentate discharges or the time to occurrence of the first large behavioral seizure. This seems somewhat unlikely, and would suggest a special or privileged input, one which was especially able to drive expression of seizures and epileptiform activity.
The other option, which would also be surprising, is that no external input is necessary. In this scenario, the spontaneous dentate discharges are an internally generated event. And the ‘latent period’ in the model represents only the further changes that occur in the dentate and downstream to allow the spread of the epileptiform activity and motor involvement. This is not to say that entorhinal input is never important–even in this particular model of epilepsy the initial insult is derived from excessive perforant path stimulation. It does, however, suggest that once an initial insult creates a proepileptic hippocampus, the entorhinal cortex may (at least in some circumstances) take a backseat, and even become irrelevant.
Whether the dentate fails to gate input, or can be a direct seizure generator, one thing is clear: the dentate is critically important in temporal lobe seizures. Optogenetic stimulation of dentate granule cells is sufficient to induce seizures, and on-demand optogenetic inhibition of granule cells is sufficient to stop spontaneous seizures in the intrahippocampal kainate model (6). Similarly, activation of hippocampal inhibitory inter-neurons is able to inhibit both spontaneous seizures in chronically epileptic animals (7) and acute kainate-induced seizures (8). In contrast, and fitting with the findings reported by Meyer et al., activation of entorhinal cortex interneurons produces no significant effect on acute seizure activity in the dentate (8).
To some extent, these findings may come of little surprise. The hippocampus is already recognized as a site of onset for temporal lobe seizures (9, 10). The modification in thinking may be only that the onset requires no external kick start, or the source of that external input. In considering how seizures may start, it is informative and truly remarkable how little of a difference perforant path transection had on ictogenesis in the recent study by Meyer et al.
Editor's Note: Authors have a Conflict of Interest disclosure which is posted under the Supplemental Materials link.