Dissecting the Fatal Seizure: The Importance of Forebrain Neurons in SUDEP

Abstract

Seizures and Premature Death in Mice with Targeted Kv1.1 Deficiency in Corticolimbic Circuits

Kelsey Paulhus, Edward Glasscock, Brain Commun. 2025; 7(1:fcae444. https://doi.org/10.1093/braincomms/fcae444

Sudden unexpected death in epilepsy (SUDEP) is the leading cause of epilepsy-related death, likely stemming from seizure activity disrupting vital brain centers controlling heart and breathing function. However, understanding of SUDEP's anatomical basis and mechanisms remains limited, hampering risk evaluation and prevention strategies. Prior studies using a neuron-specific Kcna1 conditional knockout mouse model of SUDEP identified the primary importance of brain-driven mechanisms contributing to sudden death and cardiorespiratory dysregulation; yet, the underlying neurocircuits have not been identified. Using the Emx1-Cre driver, we generated a new conditional knockout mouse model lacking Kcna1 in excitatory neurons of the cortex, hippocampus, amygdala, and select vagal afferents. To test whether the absence of Kv1.1 in forebrain corticolimbic circuits is sufficient to induce spontaneous seizures, premature mortality and cardiorespiratory dysfunction, we performed survival studies and electroencephalogram, electrocardiogram, and plethysmography (EEG-ECG-Pleth) recordings. We demonstrate premature death and epilepsy in corticolimbic conditional knockout mice. During monitoring, we fortuitously captured one SUDEP event, which showed a generalized tonic-clonic seizure that initiated respiratory dysfunction culminating in cardiorespiratory failure. In addition, we observed that cardiorespiratory abnormalities are common during nonfatal seizures in conditional knockout mice, but mostly absent during interictal periods, implying ictal, not interictal, cardiorespiratory impairment as a more reliable indicator of SUDEP risk. These results point to corticolimbic excitatory neurons as critical neural substrates in SUDEP and affirm seizure-related respiratory and cardiac failure as a likely cause of death.

Commentary

Sudden unexpected death in epilepsy (SUDEP) is the unexpected and unexplained cause of death of a person with epilepsy. SUDEP is the most common cause of epilepsy-related mortality, accounting for 8% to 17% of all epilepsy-related deaths, increasing to 50% for those patients refractory to treatment. ¹ SUDEP is second only to stroke for number of life-years lost. ² Despite its significant impact, the pathophysiological underpinnings of SUDEP remain incompletely understood, in part due to the unpredictability and rarity of the event, which poses a challenge for clinical study. The most impactful clinical study so far is over a decade old and involves observations in 9 patients who died from fatal seizures while in epilepsy monitoring units (EMUs). This MORTEMUS (MORTality in EMUS) study demonstrated that all cases of SUDEP observed in EMUs resulted from respiratory arrest that occurred secondary to a seizure. ³

Repeated generalized tonic-clonic seizures (GTCSs) are the single most significant risk factor for SUDEP. ² Although refractory GTCSs are the number one risk factor, examples abound of SUDEP cases in patients who experienced very few seizures and appeared to have them under control. Probably the most famous example is that of Cameron Boyce, the actor, who died after having only four documented seizures in his lifetime. Of course, whether Cameron's final seizure was a GTCS is not known. However, this does not exonerate GTCSs. All fatal seizures in MORTEMUS were classified as tonic-clonic. In addition, to date, all fatal seizures in mice would similarly fit this diagnosis. These 2 points underscore the apparent necessity of GTCSs, at least for the most understood cases of SUDEP. However, many patients experience hundreds or even thousands of GTSCs and live, implying that something in the manifestation of certain GTCS must impair breathing.

The question is then, what neural circuits drive tonic-clonic seizures to be fatal?

This is precisely the question that Paulhus and colleagues set out to study as described in their recent publication in Brain Communications. ⁴ The authors use a Kcna1 conditional KO (cKO) mouse. Germline Kcna1 KO mice have become a standard genetic epilepsy model used in preclinical SUDEP research, in addition to the use of models of Dravet Syndrome, SCN8A epilepsy, and other epilepsies that result from ion channel mutations.^5,6 Both gain and loss-of-function mutations in KCNA1 can cause epilepsy in humans. Many KCNA1 variants identified are associated with severe developmental and epileptic encephalopathy and other epilepsy disorders. ⁷

In the current study, the authors created a Kcna1 cKO mouse, where knock out was restricted to excitatory neurons of the forebrain (cerebral cortex, hippocampus, and amygdala) by crossing Emx1-Cre^+/– mice with Kcna1^flox/− mice. The Kcna1 cKO mice exhibited spontaneous seizures of various convulsive phenotypes, with many having the hallmarks of GTCSs. The mice also had a survival curve where about 30% died by 100 days of age. For comparison, at this time point, there was about 50% and 80% mortality in mice where Kcna1 was knocked out of all neurons and all cells, respectively. ⁸ Similarly, when gain-of-function Scn8a mutations were knocked into forebrain neurons selectively, this prolonged average lifespan but still resulted in 100% mortality compared to germline KO. ⁹ This suggests alteration, presumably hyperexcitability, of brainstem neurons can contribute to a more severe SUDEP phenotype; however, it is unclear whether this implicates brainstem neurons participating in driving a fatal seizure, or a secondary effect like impaired cardiorespiratory homeostasis.

Paulhus and colleagues also instrumented a subset of the Kcna1 cKO mice for cardiac and respiratory measurement during seizures and found that some seizures caused cardiac arrhythmias and respiratory dysfunction. Cardiac dysfunction was rare, occurring in less than 27% of seizures, and included conduction block, tachycardia, and bradycardia. Using chronic whole-body plethysmography recordings, a much greater proportion (up to 70%) of seizures were found to produce respiratory dysfunction. The most common respiratory phenotype was tachypnea or increased respiratory rate. Considering the energetic demands of seizures, an increased breathing rate may not be unexpected; however, suppressed breathing in the form of hypopnea, bradypnea, and apnea was also observed in many seizures. The authors also point out that cardiac dysfunction never preceded, and usually (about 80%) followed, respiratory dysfunction. This finding is significant as it suggests that cardiac anomalies may result from prior dysfunctional breathing, as was observed in all SUDEP cases in the MORTEMUS study. ³

The authors were fortunate to catch a single fatal seizure with adequate cardiorespiratory monitoring. The fatal seizure was a GTCS with an extended tonic phase at the end that coincided with bradycardia and apnea. This is similar to observations in germline KO of Kcna1 and other models.^5,6,10 Prior to the fatal seizure, this mouse had experienced 15 GTCS-equivalent seizures in the preceding 24 h. In addition, this mouse had the longest seizure durations of any mouse recorded in the study. Taken together, the mouse that died from SUDEP had a significant seizure burden that may have led to the fatality.

As mentioned above, a dichotomy exists in SUDEP occurrence. On the one hand, the number of GTCSs is a clear predictor of SUDEP likelihood. On the other hand, numerous cases of SUDEP where few seizures have been observed and/or seizures are well controlled are reported. The current preclinical study provides a model of the former situation. The single observed fatality was in a mouse that had a high seizure burden. A question going forward: “Is seizure burden a correlation or a cause?” That is, did the repetitive seizures in the Kcna1 cKO mouse cause lasting changes (ie, plasticity) that enabled production of the eventual terminal seizure? Or is a more severe seizure phenotype also results in fatality? Perhaps, based on the clinical data, the circumstances of a fatal seizure are somewhat stochastic and a larger sample set of seizures simply increases the odds. More work is needed, but a role for the cortical and limbic system cannot be ignored when trying to understand the neural circuitry of SUDEP.

We still know so little about SUDEP. Identifying biomarkers and therapeutically relevant mechanisms will be critical in the development of treatment strategies. It may be that patients with prolonged and frequent seizures, and possibly co-occurring cardiorespiratory dysfunction, are most at risk and would benefit from intervention. Having a diverse range of animal models that explore different pathways to the same outcome will help us identify other risk factors and lead to better outcomes for patients with epilepsy.