That Sync-ing Feeling: The Progressive Drop of Hippocampal and Entorhinal Coordination in Murine Epilepsy

Abstract

Distinct Changes to Hippocampal and Medial Entorhinal Circuits Emerge Across the Progression of Cognitive Deficits in Epilepsy

Feng Y, Diego KS, Dong Z, Christenson Wick Z, Page-Harley L, Page-Harley V, Schmipper J, Lamsifer SI, Pennington ZT, Vetere LM, Philipsberg PA, Soler I, Jurkowski A, Rosado CJ, Khan NN, Cai DJ, Shuman T. Cell Rep. 2025;44(2):115131.

Temporal lobe epilepsy (TLE) causes pervasive and progressive memory impairments, yet the specific circuit changes that drive these deficits remain unclear. To investigate how hippocampal-entorhinal dysfunction contributes to progressive memory deficits in epilepsy, we performed simultaneous in vivo electrophysiology in the hippocampus (HPC) and medial entorhinal cortex (MEC) of control and epileptic mice 3 or 8 weeks after pilocarpine-induced status epilepticus (Pilo-SE). We found that HPC synchronization deficits (including reduced theta power, coherence, and altered interneuron spike timing) emerged within 3 weeks of Pilo-SE, aligning with early-onset, relatively subtle memory deficits. In contrast, abnormal synchronization within the MEC and between HPC and MEC emerged later, by 8 weeks after Pilo-SE, when spatial memory impairment was more severe. Furthermore, a distinct subpopulation of MEC layer 3 excitatory neurons (active at theta troughs) was specifically impaired in epileptic mice. Together, these findings suggest that hippocampal-entorhinal circuit dysfunction accumulates and shifts as cognitive impairment progresses in TLE.

Commentary

Temporal lobe epilepsy (TLE) exerts a significant burden on the quality of life of those with the disease. Beyond seizures, many patients show comorbid cognitive and neuropsychiatric symptoms, which are often poorly managed by conventional therapy. ¹ These symptoms represent a significant unmet need in the clinic, and accordingly, there is a substantial knowledge gap surrounding the pathophysiology of these aspects in epilepsy syndromes. Using the pilocarpine mouse model of TLE, in which systemic pilocarpine triggers a period of status epilepticus (SE), Feng et al ² combined behavior and electrophysiology to investigate the underexplored neural basis of cognitive sequelae of epileptogenesis. They show that a progressive worsening of circuit disruption between the hippocampus (HPC) and medial entorhinal cortex (MEC) accompanies increasingly severe behavioral impairments.

The authors used multiple time points to show that spatial memory deficits in the pilocarpine TLE model worsen progressively between 3 and 8 weeks post-SE. They employed a novel object location task with two “difficulty” levels: acclimatization to the objects for either 2 (hard) or 3 (easy) 6-min trials prior to testing. Control mice were able to distinguish between the moved object and the nonmoved object at both difficulty levels. However, epileptic mice failed to distinguish between the objects at 8-weeks post-SE in the easy task, and at both time points in the hard task. Interestingly, this behavioral deficit was dissociated from both seizure and interictal discharge frequency, and was not linked to cell death, as more cell death was apparent 2 days post-SE compared to 3 and 8 weeks afterwards. ² This suggests that the pilocarpine epilepsy model produces a progressive cognitive deficit distinct from the acute effect of seizures and SE. One potential confound is the short acclimatization period, where epileptic mice in particular may be more neophobic and anxious, and therefore may not explore the arena or objects enough to learn their relative position or may prefer familiarity. ³ This may explain the high variance in the behavioral data, where mice may not have had the opportunity to learn which object has moved, or may avoid novelty altogether. Another potentially interesting analysis would have been to correlate the time from the last seizure with cognitive performance.

Next, the authors turned to in vivo acute electrophysiological recordings conducted simultaneously in the HPC and MEC during a virtual reality linear track task. ² Initially, they analyzed local field theta (5–12 Hz) oscillations in these regions, and found that theta power was reduced in epileptic animals in specific layers of CA1 and dentate gyrus (DG), with the power reduction being greater in 8-week than 3-week post-SE animals, consistent with the authors’ previous work. ⁴ They also found deficits in theta phase coherence (a measure of how consistently aligned oscillations are) within HPC subregions: at 3 weeks post-SE, cross-coherence was significantly reduced between the hilar region of the DG (DGhil) and much of the HPC. At 8 weeks, further desynchronization was observed between DGhil and the CA1 pyramidal cell layer. ² The authors then considered that circuit pathology in MEC may precede HPC pathology. Indeed, the HPC theta rhythm is believed to be driven (in part) by MEC, and optogenetic inhibition of MEC disrupts HPC theta activity in subregions where theta power is most reduced in epilepsy. ⁵ However, the authors found no disruption of phase coherence between the MEC and HPC at 3 weeks post-SE. Instead, significant reductions in coherence were only seen between MEC layer 1 and HPC subregions at 8 weeks.

In their previous work, the authors found a breakdown in place cell spatial information content and a reduction in place field stability in the pilocarpine epilepsy model. ⁴ This instability was only evident after 5–6 weeks post-SE, matching the timeline of desynchronization between MEC and HPC. As MEC neurons project strongly to HPC subregions and can drive their theta rhythm, the authors examined MEC excitatory neurons and clustered 2 populations in layer 3 (L3) based on their theta modulation depth and preferred phase. ¹ Cells firing at the peak of theta displayed low modulation depth and little change over the development of epilepsy. However, those with preferential firing at the trough of theta displayed high theta entrainment in control animals, and a stark reduction in modulation depth at 8 weeks post-SE. These neurons therefore appear to be more sensitive to the pathology induced by epilepsy.

The discovery of a functionally distinct excitatory neuron population in MEC L3 is particularly novel, as L3 has previously been considered homogeneous with only weak theta-phase locking in its neurons. ⁶ Until now, no distinct subpopulation has been observed. ⁷ While it is unclear why this population has not been observed before, these neurons may constitute a novel route for information to be propagated into the HPC and warrant further characterization both in health and in disease, including for spatial modulation. Hippocampal CA1 receives strong input directly from MEC L3, and this monosynaptic pathway has been implicated in hippocampal-dependent memory, ⁸ such as that required for the object-location task. Current models of hippocampal theta consider multiple rhythm generators to exist, ⁹ and changes in subpopulations within MEC L3 may indicate the disruption of 1 of these generator circuits emerging between 3 and 8 weeks post-SE.

The use of the pilocarpine epilepsy model could be a potential confound here. Pilocarpine is administered systemically and therefore may affect multiple theta-generating or theta-modulating regions beyond the MEC and HPC, including the medial septum, prefrontal and cingulate cortices, retrosplenial cortex, thalamic nuclei, and brainstem neuromodulatory centers. Considering this as a brain-wide pathology, characterizing causal discoordination between MEC and HPC is challenging. A more direct approach, such as focal epileptogenesis and putatively therapeutic manipulations bidirectionally modulating spatial behavior, may provide clearer evidence for the causal relationship between the cognitive sequelae of epileptogenesis and HPC-MEC circuit pathology.

Overall, the relationship between the emergence of spatial cognitive deficits and (para)hippocampal desynchrony is compelling. Feng et al ² findings shed light on the longitudinal cognitive decline in TLE, the timing of alterations in different brain regions, and the properties of theta-oscillations associated with TLE progression. These findings help in understanding the pathophysiology and epileptogenesis of TLE and have implications for potential treatments of comorbidities in epilepsy. Moving forward, more work is needed to understand the underlying processes of epileptogenesis in TLE, to assess longitudinal cognitive decline also in female animals, and more complex spatial memory behavior paradigms in freely moving animals. This could be the next step to decipher this important and understudied phenomenon.