Seizing Control of the CA2

Abstract

Chemogenetic Activation of Hippocampal Area CA2 Promotes Acute and Chronic Seizures in a Mouse Model of Epilepsy

LaFrancois JJ, Kennedy M, Rathod M, Santoro B, Lisgaras CP, Siegelbaum SA, Scharfman HE. Neurobiology of Disease. 2026;219:107256.

Pyramidal cells (PCs) of hippocampal area CA2 exhibit increased excitability in temporal lobe epilepsy (TLE) and in mouse models of TLE. In epileptic mice, selective inhibition of CA2 PCs reduces chronic seizures. Here we asked if activating CA2 PCs increases seizures. Mice expressing Cre recombinase in CA2 PCs (Amigo2-Cre mice) were injected with the convulsant pilocarpine to induce a period of severe seizures (status epilepticus, SE), which leads to chronic seizures after 3–4 weeks (epilepsy). Epileptic mice were injected with a Cre-dependent adeno-associated virus (AAV) to express an excitatory designer receptor exclusively activated by designer drug (eDREADD; hM3Dq) in dorsal CA2 bilaterally and implanted with subdural EEG electrodes. After recovery, mice were recorded continuously using video and EEG for 6 weeks, 3 weeks with drinking water containing the eDREADD activator clozapine-N-oxide (CNO) and 3 weeks without CNO. CA2 activation with CNO caused a significant increase in seizure frequency and duration. Seizures occurred in clusters (many seizures per day over several consecutive days) and mice given water with CNO had a greater maximum number of seizures per day during a cluster compared to water without CNO. CNO had no significant effect in control mice. In naïve Amigo2- Cre mice expressing hM3Dq, pre-treatment with CNO before pilocarpine administration shortened the latency to SE and increased EEG power at the start of SE. Taken together with prior findings, the results suggest that CA2 is a control point for regulating seizures in the pilocarpine mouse model of TLE.

Commentary

The hippocampus plays a key role in many processes by regulating the balance between excitation and inhibition across its various regions, as well as its incoming and outgoing connections. Classically, the trisynaptic circuit consists of information passing through the entorhinal cortex → dentate gyrus (DG) → CA3 → CA1 regions. Until recently, the CA2 region of the hippocampus was understudied, and its connections and function were largely unknown. Interest in the CA2 region began to increase when postmortem human studies noted alterations of the CA2 region in several neurological disorders, including accumulation of α-synuclein in Parkinson's disease, loss of parvalbumin interneurons in schizophrenia, and accumulation of tau pathology in Alzheimer's disease. ¹ However, the CA2 region is resistant to neuronal death in temporal lobe epilepsy (TLE) in both humans and rodent models.^2,3

The recent development of tools such as transgenic mouse lines, viral vectors, and optogenetics has allowed for the study of CA2 connections and function. The CA2 region receives excitatory input from layer 2 of the entorhinal cortex, CA3 pyramidal cells, and DG granule cells.^4–6 However, it also receives long-range input from other structures, including supramammillary nuclei, medial septum, and median raphe nucleus.^4,6 Interestingly, CA2 pyramidal neurons can bilaterally project to CA1, CA2, and CA3 regions. ⁶ In addition to the classical trisynaptic circuit, a powerful disynaptic circuit between the entorhinal cortex → CA2 → CA1 was identified. ⁷ The Amigo2-Cre transgenic mouse line, developed by Hitti and Siegelbaum, ⁶ expresses Cre recombinase primarily in CA2 pyramidal cells with some expression outside of the CA2 in the adult mouse. Using the Amigo2-Cre mouse, it was discovered that inactivation of CA2 pyramidal cells results in profound deficits in social recognition memory, but not sociability or other aspects of hippocampal-dependent memory such as spatial memory. ⁶ Consistent with these observations, excitotoxic lesion of the CA2 in rats also resulted in social memory deficits. ⁸ Together, these observations highlight the complexity of CA2 connections and underscore the important role that it plays in balancing neuronal excitation and inhibition and its contribution to processes such as social memory.

Early studies on hippocampal circuitry in epilepsy drew attention to the CA2 region because of its distinctive resilience. Unlike the adjacent areas, the CA2 region is largely spared from neuronal death in both human patients and rodent models of TLE.^2,3 CA2 pyramidal cells from TLE patients show increased excitability ⁹ ; however, it was unknown whether suppressing activity of CA2 pyramidal cells would be beneficial. Indeed, Whitebirch et al ¹⁰ showed that inhibition of CA2 pyramidal cells in the pilocarpine mouse model of mesial TLE (MTLE) significantly reduced spontaneous seizure frequency. Thus, these findings predict that increasing the activity of CA2 pyramidal cells would worsen seizure outcomes.

In the current manuscript, LaFrancois et al ¹¹ investigated the effects of CA2 pyramidal cell activation on acute seizures and chronic epilepsy. To examine the effect of CA2 activation on acute seizures, Amigo2-Cre mice were injected with a Cre-dependent adeno-associated virus (AAV) expressing eDREADD hM3dq tagged with mCherry (AAV-hM3Dq-mCherry) in the CA2. Four weeks later, mice were administered clozapine-N-oxide (CNO) 30 min prior to administration of the proconvulsant pilocarpine. The authors found that activation of CA2 resulted in a significant increase in EEG power and shorter latency to status epilepticus (SE), demonstrating that the activation of CA2 increases susceptibility to acutely induced seizures. To examine the effect of CA2 activation on chronic epilepsy, 8-week old Amigo2-Cre mice were subjected to SE via the administration of pilocarpine, and after 4 weeks, EEG electrodes were implanted and AAV-hM3Dq-mCherry was bilaterally injected into the dorsal CA2. Four weeks later, continuous EEG recordings were obtained for 6 weeks. Mice received water with or without CNO for a 3-week period followed by the converse treatment for another 3 weeks, allowing for within animal controls. The authors found that MTLE mice that were Cre+ and injected with AAV-hM3Dq-mCherry exhibited higher frequencies of spontaneous seizures when administered CNO in the drinking water (from an average of 39.5 to 72.5 seizures over a 3-week period), indicating that activation of CA2 increases seizure frequency. Interestingly, this effect was only observed in male mice, although the greater variability of seizure frequency in female mice may have limited the ability to detect a comparable effect. The authors also noted that CA2 activation resulted in significantly longer seizure duration and an increase in the number of seizures in a seizure cluster (defined as ≥2 seizures/day for ≥2 consecutive days). Seizure severity, the time of day of each seizure, and EEG power at different frequency bands were not altered by CA2 activation. Taken together, LaFrancois et al ¹¹ demonstrate that the CA2 hippocampal region facilitates seizure susceptibility and activation of the CA2 exacerbates susceptibility to acute seizures and chronic epilepsy in a mouse model of MTLE.

Given the complex projections of CA2 pyramidal neurons in the epileptic brain, ¹⁰ it is not surprising that CA2 activation increased susceptibility to acutely induced seizures and chronic epilepsy in an MTLE model. In the pilocarpine MTLE model, there is increased excitatory output to the CA1 region and GABAergic inhibition is reduced, ¹⁰ so further increasing excitability of the CA2 likely shifts the balance to more excitation and less inhibition. Interestingly, LaFrancois et al ¹¹ found no correlation between seizure frequency and seizure duration, or seizure duration and cluster peaks (the maximum number of seizures in a cluster), suggesting that these features might be regulated by different underlying mechanisms. For instance, an increase in seizure frequency with CA2 activation might reflect synchronization of neural activity in downstream CA1, whereas prolonged seizure duration could arise from sustained, repetitive firing between the hippocampus and cortex driven by increased excitatory output between CA2 and CA1. Previous reports demonstrate that the CA2 region is spared from neuronal death in MTLE models, and this was also confirmed in the current study.^3,10,11 However, it represents a missed opportunity to determine whether the CA2 region remains spared from neuronal loss following excitation of the CA2 region in the pilocarpine MTLE model, despite more frequent spontaneous seizures. This observation might provide insight into whether seizure susceptibility and resistance to neuron loss in the CA2 are regulated by distinct mechanisms. Additionally, given the critical role of the CA2 in social memory, it would be valuable to investigate whether its activation in MTLE affects social memory. Together, these observations highlight the importance of CA2 in modulating seizure dynamics in MTLE, while underscoring key gaps in our understanding of the impact of CA2 on neuronal integrity and behavior in MTLE, and more broadly, the potential role of the CA2 in other forms of epilepsy.