Abstract
Optogenetic Dissection of Ictal Propagation in the Hippocampal–Entorhinal Cortex Structures
Lu Y, Zhong C, Wang L, Wei P, He W, Huang K, Zhang Y, Zhan Y, Feng G, Wang L. Nat Commun 2016;7:12019. doi:10.1038/ncomms12019.
Temporal lobe epilepsy (TLE) is one of the most common drug-resistant forms of epilepsy in adults and usually originates in the hippocampal formations. However, both the network mechanisms that support the seizure spread and the exact directions of ictal propagation remain largely unknown. Here we report the dissection of ictal propagation in the hippocampal–entorhinal cortex (HP–EC) structures using optogenetic methods in multiple brain regions of a kainic acid-induced model of TLE in VGAT-ChR2 transgenic mice. We perform highly temporally precise cross-area analyses of epileptic neuronal networks and find a feed-forward propagation pathway of ictal discharges from the dentate gyrus/hilus (DGH) to the medial entorhinal cortex, instead of a reentrant loop. We also demonstrate that activating DGH GABAergic interneurons can significantly inhibit the spread of ictal seizures and largely rescue behavioural deficits in kainate-exposed animals. These findings may shed light on future therapeutic treatments of TLE.
Commentary
Temporal lobe epilepsy (TLE) is one of the most common seizure disorders, accounting for about 60% of all cases of epilepsy (1). To develop novel therapeutic interventions for TLE, we must understand how neuronal circuits within the temporal lobe drive seizure activity. This is no simple task, however, as the temporal lobe contains multiple highly interconnected circuits (2). Separating activity originating in one region from activity originating in another requires precise temporal and spatial measurement of neuronal activity. Fortunately, advances in both in vivo recording techniques (3) and optogenic cell-type specific stimulation (4) have provided exciting new insights into normal and pathological circuit activity in the temporal lobe.
The hippocampus, subiculum, entorhinal cortex, and perirhinal cortex comprise the temporal lobe. In the hippocampus, dentate granule cells receive excitatory input from the entorhinal cortex. The dentate gyrus (DG) is often thought of as an “inhibitory gate” that filters incoming activity and allows only salient activity to propagate through the circuit (5). From the DG, mossy fibers project to area CA3 pyramidal neurons. CA3 neurons are highly interconnected with each other, allowing them to amplify incoming activity from the DG. CA3 neurons then send excitatory outputs to area CA1, a region often associated with sclerotic damage in the epileptic brain. Excitatory axons from CA1 project back to the entorhinal cortex, completing the “reentrant loop.” Excitation at each step is carefully controlled by powerful inhibition mediated by a variety of GABAergic interneurons (6). As one might imagine, a reentrant circuit made up of powerful feed-forward excitation can easily be prone to generating unconstrained, hyper-synchronized activity when disrupted.
Many studies using hippocampal–entorhinal cortex in vitro brain slices suggest that ictal activity, evoked by a variety of manipulations, begins in the entorhinal cortex (EC), propagates first to the DG and then into hippocampal areas CA3 and CA1 before feeding back to the EC (7, 8). A recent study, however, elegantly shows that ictal activity can propagate in the reverse direction, from the DG to the EC, in vivo. Furthermore, the authors went on to demonstrate that attenuating ictal activity in the DG eliminated ictal activity in the EC, but not vice versa. Lastly, they showed that optogenic activation of GABAergic interneurons in the DG aborted behavioral and electrographic seizures. This is an important finding that requires an adjustment of how we envision temporal lobe circuit activity during seizures, with a few study-specific caveats.
To perform these studies, Lu and colleagues constructed novel neural probes for simultaneous multi-site optical stimulation and electrophysiological recording in vivo. Two probes were implanted, one into the DG and one into the EC. At the same time, a microinfusion tube was implanted to allow the local infusion of the convulsant kainic acid (KA) into the dorsal hippocampus. KA was then infused unilaterally, and ictal discharges began within approximately 30 minutes. Ictal activity was characterized by large-amplitude spikes and multi-unit bursts indicative of neuronal hyper-synchronization. The authors utilized cross-area spike-triggered averages of the local field potentials to examine the flow of activity through the temporal lobe. Put simply, this is a way to examine the sequential activation of connected brain regions based on the time it takes for a signal to propagate from one region to another. Instead of the reentrant loop seen in many studies, the authors found that ictal activity propagated from the DG to the EC.
To determine if activity in the DG did, in fact, elicit ictal activity in the EC, the authors utilized mice genetically engineered to allow light-induced activation of GABAergic interneurons. Using this optogenetic approach, the authors showed that in vivo activation of GABAergic interneurons suppressed local ictal activity in both the DG and the EC. Of exciting import, optogenetic activation of interneurons in the DG suppressed ictal propagation to the EC. Equivalent activation of GABAergic interneurons in the EC, however, did not suppress ictal activity in the DG. This suggests that the ictal activity in the DG contributes to ictal activity in the EC, but not vice versa. Additional studies suggest that activity in the beta and gamma bands may be especially important for DG to EC ictal propagation. Finally, cyclic optical stimulation (130 Hz, 5 milliseconds, 1 min/cycle, 5-minute intercycle interval) of GABAergic interneurons in the DG attenuated seizure activity after 30 minutes of stimulation and reduced behavioral seizures. These studies suggest that the DG may be more than an “inhibitory gate” but may also actively generate ictal activity that propagates to the EC and through the brain.
To set these findings into context, one must first consider the seizure model used by Lu and colleagues. Infusion of KA into the dorsal hippocampus mimics TLE in its focality but likely oversimplifies how seizures initiate. In systemic convulsant models of chronic epilepsy, elegant multi-site in vivo recording suggests that more dynamic and diverse circuits can initiate spontaneous seizures (9, 10). Work from the Buckmaster lab shows that ictal and preictal activity begin throughout the temporal lobe, propagate locally, and somewhat stochastically, before generalizing(9, 10). Next, the exciting finding that optogenetic stimulation of GABAergic interneurons in the DG can suppress seizures should also be carefully considered. Recent studies show that seizures can be aborted via optogenetic interventions in the thalamus (11), hippocampus (12), superior colliculus (13), and cerebellum (14). How long-range projections from activated neurons affect circuit activity and the specificity of channelrhodopsin expression and light delivery should always be considered in interpreting these types of studies. In another intriguing recent study, optogenetic inactivation of DG granule cells stopped seizures (15). Does this manipulation suppress DG activity that was propagating to the EC as Lu and colleagues’ study suggests? Lastly, human temporal lobe data clearly implicates the EC in the pathology of TLE (16). The majority of patients with medically intractable TLE have reduced EC volume, which correlates with hippocampal tissue loss and the duration of epilepsy (17). Even though the study by Lu and colleagues demonstrates ictal propagation from DG to EC, the EC undoubtedly plays an important role in the progression of human epilepsy.
With these considerations in mind, Lu and colleagues work provides a new pathway to consider when dissecting ictal activity in the temporal lobe. Their work clearly shows that ictal activity generated in the DG can propagate to the EC, the opposite of how we normally consider activity to flow through the “reentrant loop” of the temporal lobe. This study highlights how novel technologic advances in in vivo recording strategies and genetic manipulation of neuronal activity with cell-type specificity can aid our understanding of the pathogenesis of epilepsy.