To Sleep,Perchance to Seize . Ay,There's the Rub

Abstract

Preoptic Area Influences Sleep-Related Seizures in a Genetic Epilepsy Mouse Model.

Potesta CV, Cargile MS, Yan A, Xiong S, Macdonald RL, Gallagher MJ, Zhou C. Cerebral Cortex. 2025 July;35(7):bhaf187. https://doi.org/10.1093/cercor/bhaf187

In patients with refractory epilepsy, states of sleep and wakefulness affect the expression of seizures. However, the mechanism by which subcortical sleep circuitry affects seizures is unknown. Here, using Gabrg2^Q390X knock-in (KI) genetic epileptic mouse model, we found that during sleep, subcortical preoptic area (POA) neurons were active in het Gabrg2^Q390X KI mice and their activity preceded or/and coincided with epileptic (poly)spike–wave discharges. Optogenetic manipulating the POA activity altered sleep/wake periods in wild-type (wt) and the het Gabrg2^Q390X KI mice. Most importantly, short-period optogenetic activation of epileptic cortical neurons alone did not effectively trigger seizures in the het Gabrg2^Q390X KI mice, while optogenetic activation of the POA nucleus slightly influenced spontaneous epileptic activity in the het Gabrg2^Q390X KI mice. In contrast, coordinated optogenetic activation/suppression of the subcortical POA nucleus with the optogenetic activation of epileptic cortical neurons effectively enhanced or suppressed epileptic activity in the het Gabrg2^Q390X KI mice, indicating that the subcortical POA activation exacerbates seizures in the het Gabrg2^Q390X KI mice. In addition, suppression of the subcortical POA nucleus decreased myoclonic jerks in the Gabrg2^Q390X KI mice. Overall, this study reveals a circuit-based mechanism of sleep-preferential seizures in one genetic epilepsy model with implications for refractory epilepsy.

Commentary

Macbeth observes, “Innocent sleep. Sleep that soothes away all our worries. Sleep that puts each day to rest. Sleep that relieves the weary laborer and heals hurt minds. Sleep, the main course in life's feast, and the most nourishing.” And indeed, for most people, sleep is a creature comfort, a respite from the struggles of life. However, for some with epilepsy, it does not provide relief from their seizures, and in many cases can exacerbate them.

Nocturnal seizures occur during sleep and are pronounced in several types of epilepsy. Nocturnal seizures can be difficult to diagnose and, of course, go underreported. However, they are of great concern. Nocturnal seizures disrupt sleep, resulting in daytime drowsiness. They can cause injury and even death. There is increased risk of sudden death (SUDEP) for those with a history of nocturnal seizures, ¹ perhaps due to the more severe hypoxia and postictal generalized EEG suppression experienced during nocturnal seizures. ²

Seizures are commonly understood to be hyperactivity of the brain. Why then would sleep, when the brain is resting, produce seizures? Of course, the neuroscientists and neurologists in the room know this premise is false. Our brains are highly active during sleep. During sleep, the brain performs essential functions, including memory consolidation. Everyone who plays an instrument knows that motor memory needs a good night's sleep. Thus, brain metabolism often increases during certain sleep stages, like rapid eye movement (REM) sleep, and only drops by 15% during certain phases of non-REM (NREM) sleep. ³

Another questionable premise is that seizures are a product of hyperactivity, when what we observe is actually hypersynchronous neuronal activity. Sleep also produces phases of rhythmic EEG activity due to neuronal synchronization. Classically, we think of delta waves (1-4 Hz) during NREM and theta waves (4-8 Hz) during REM, which both result from synchronous neural activity. More subtle, but significant, EEG manifestations during sleep include sleep spindles and K complexes. These phenomena have essential roles in memory consolidation during sleep. ⁴ However, their forms are remarkably similar to EEG signals observed during seizures. Sleep spindle neural circuitry has been proposed to be hijacked to create spike–wave discharges (SWDs) that occur during many types of seizures, ⁵ and K complexes have been shown to trigger SWDs during the transition from NREM and REM sleep. ⁶

The neural circuitry of the thalamus and cortex is thought to underlie both sleep spindles and K complexes. However, these thalamocortical circuits are not implicated in creating or maintaining sleep states. The neural substrates that drive sleep are subcortical, starting in the hypothalamus and extending down into the medulla. ⁷ The preoptic area (POA) in the hypothalamus is a key region that drives sleep. Lesioning of the POA reduces sleep time, and, more recently, optogenetic and chemogenetic stimulation demonstrates increased sleep. ⁷ Other subcortical regions, like the parafacial zone and ventrolateral periaqueductal gray matter, influence sleep. However, if and how these sleep-promoting neurons influence seizures is largely unstudied.

This was the question Potesta and colleagues set out to answer in their recent paper. ⁸ The authors of this study set out to investigate a novel hypothesis—that subcortical structures, such as the POA, might also play a role in driving nocturnal seizures. This hypothesis was based on their previous work with a mouse model of Dravet syndrome, which showed increased seizure incidence during sleep.. ⁹ This mouse model harbors a loss-of-function Q390X mutation in a single Gabrg2 allele (Het Gabrg2^Q390X KI mice). All experiments involved the use of littermate wild-type control mice (Wt).

First, Potesta et al confirmed that POA activity drives sleep. The authors demonstrate that neurons of the POA are active during NREM sleep with multiunit recording in tandem with standard EEG recording. POA neurons were an order of magnitude more active during NREM than during wakefulness. Interestingly, the POA unit activity also immediately preceded many SWDs in the Het mice, suggesting the POA might drive these events as well.

To test whether the POA could drive NREM sleep, the authors used an optogenetic approach. They backcrossed the Wt and Het mice with mice that expressed both the tetracycline transactivator downstream of the cFos promoter and channelrhodopsin2 (ChR2) and Halorhodopsin (NpHR) downstream of the Tet Operator. Thus, these mice expressed both ChR2 and NpHR in neurons that express Fos (ie, activated neurons). Due to the different wavelengths of light needed to stimulate ChR2 and NpHR (blue and yellow, respectively), the authors were able to photostimulate and photoinhibit the seizure-activated neurons of the POA. As expected from previous studies, the authors found that photostimulation of the POA increased NREM and photoinhibition decreased NREM.

Next, the impact of POA photomanipulation on spontaneous seizures was assessed. Using the same methods as above, the authors photostimulated and photoinhibited the POA of Wt and Het mice while recording spontaneous SWDs. Photostimulation moderately increased both the frequency and duration of SWDs. However, photoinhibition did not have a significant impact on any seizure parameter. This suggests that although the POA may be able to stimulate seizures, its activity is not driving the spontaneous seizures observed in these mice.

Of course, it is not expected that the POA is solely responsible for seizure initiation. It is possible, even probable, that POA activity, in conjunction with other neuronal activity, may increase seizure pressure. Thus, the authors concocted a more delicate seizure paradigm by using a brief cortical stimulation in the middle of the POA photostimulation or photoinhibition. A single pulse of light blue light was applied to the somatosensory cortex (S1). In this case, POA photostimulation had a much larger effect on SWD frequency and duration than without S1 photostimulation. In addition, a significant effect was now seen with photoinhibition of the POA, demonstrating a stronger causal role for the POA in generating seizures. Of course, none of these effects were observed in Wt mice.

In summary, Potesta et al demonstrated that the POA may contribute to seizure pressure in a mouse model of Dravet syndrome. POA neurons are active during NREM sleep and proximal to SWDs. Their activation can increase SWDs; however, suppression of POA neurons had minimal impact on spontaneous SWDs. Whether this reflects a minor role for POA or a technical limitation of the approach is unclear. Other elements of the sleep neural circuitry could also be involved. In addition, what neural circuitry connects the POA to the thalamocortical pathways that create SWDs is an interesting question, no less because the sleep-promoting neurons of the POA are thought to be inhibitory. ⁷

There is a great need to understand how seizures are generated. Understanding common seizure flashpoints might allow for novel therapeutic interventions. Nocturnal seizures are of particular concern as they often go unnoticed and can even be deadly. ¹ The current study may be the trailhead to find targets for treating nocturnal seizures, so that those currently suffering nocturnal seizures may join Henry IV in exclaiming, “O sleep! O gentle sleep! Nature's soft nurse.”