Neuronal Physiology of Generalized Seizures: The 4 Horsemen of Absence Epilepsy

Abstract

Decreased but Diverse Activity of Cortical and Thalamic Neurons in Consciousness-Impairing Rodent Absence Seizures

McCafferty C, Gruenbaum BF, Tung R, Li JJ, Zheng X, Salvino P, Vincent P, Kratochvil Z, Ryu JH, Khalaf A, Swift K, Akbari R, Islam W, Antwi P, Johnson EA, Vitkovskiy P, Sampognaro J, Freedman IG, Kundishora A, Depaulis A, David F, Crunelli V, Sanganahalli BG, Herman P, Hyder F, Blumenfeld H. Nat Commun. 2023;14(1):117. doi:10.1038/s41467-022-35535-4

Absence seizures are brief episodes of impaired consciousness, behavioral arrest, and unresponsiveness, with yet-unknown neuronal mechanisms. Here we report that an awake female rat model recapitulates the behavioral, electroencephalographic, and cortical functional magnetic resonance imaging characteristics of human absence seizures. Neuronally, seizures feature overall decreased but rhythmic firing of neurons in cortex and thalamus. Individual cortical and thalamic neurons express one of four distinct patterns of seizure-associated activity, one of which causes a transient initial peak in overall firing at seizure onset, and another which drives sustained decreases in overall firing. 40-60 s before seizure onset there begins a decline in low frequency electroencephalographic activity, neuronal firing, and behavior, but an increase in higher frequency electroencephalography and rhythmicity of neuronal firing. Our findings demonstrate that prolonged brain state changes precede consciousness-impairing seizures, and that during seizures distinct functional groups of cortical and thalamic neurons produce an overall transient firing increase followed by a sustained firing decrease, and increased rhythmicity.

Commentary

Absence seizures are a common type of generalized seizure experienced by patients with idiopathic generalized epilepsy (IGE) and are characterized by abrupt altered consciousness in conjunction with generalized spike-wave discharges (SWDs) on EEG. Although generalized seizures in IGE patients often respond to anti-seizure drugs, approximately 15% of IGE patients are resistant to appropriate medications. ¹ To develop new pharmacological and neurostimulation treatments for medically refractory generalized seizures it is critical to better understand their neurophysiology. In this paper, McCafferty et al advanced our understanding of absence seizure physiology by testing their effects on consciousness, functional magnetic resonance imaging (fMRI) signaling, and cortical/thalamic single neuron firing rates during absence seizures in genetic absence epilepsy rats from Strasbourg (GAERS). ²

For decades, absence epilepsy researchers studied inbred strains of rats and mice and, more recently, gene-targeted knockin mice that express human epilepsy mutations. ³ These rodent absence epilepsy models exhibit locomotor arrest during SWDs, but it was uncertain if this behavior cessation reflected altered consciousness. Intriguingly, a 2017 study demonstrated that rats conditioned to receive a food reward shortly after SWD termination had shorter duration SWDs than those without a reward. ⁴ Were the “absence epilepsy” rats consciously shortening their SWDs to obtain a reward, or did an experimental environment in which rewards were provided produce a state of high arousal that is naturally associated with short duration SWDs?

In order to relate consciousness in rat and human SWDs, the authors developed novel rat behavioral experiments adapted from this group’s previous study of human absence seizures. ⁵ In the previous human study, they found that absence seizures caused a greater impairment of a continuous performance task (subject shown different letters and asked to tap when shown letter “X”) than a purely repetitive task (subject tapped when shown any letter). Using GAERS, McCafferty et al found that in a continuous performance task (licking a sucrose dispensing port within 10s of a low-intensity auditory tone), the rats were substantially less likely to respond during a SWD (0.4%) than at baseline (88.2%) or immediately after a SWD (78.2%). In an analog of the repetitive task (continuous licking required to elicit an intermittent sucrose reward), GAERS exhibited a drastically lower lick rate (0.007 licks/s) during a SWD than in the 10s prior (0.75 licks/s) or immediately after (0.55 licks/s) the SWD. These behavioral studies demonstrated that SWDs produce a similar impairment of consciousness in the GAERS model as in their human counterparts.

Previous fMRI studies in IGE patients have identified brain regions that may be important in the pathogenesis of generalized seizures. Interestingly, while human and rodent models demonstrated increased thalamic blood-oxygen-level-dependent (BOLD) signal, human absence seizure patients exhibited decreased cortical BOLD signal while rodents exhibited increased cortical BOLD. A difference between rodents and humans in SWD cortical neurovascular coupling would suggest differences in their pathophysiology which may impact treatment development using rodent models. McCafferty et al hypothesized that anesthesia used for prior rodent fMRI studies could alter BOLD responses and thus devised a rat fMRI methodology that habituated the animals to the fMRI apparatus and thereby permitted the experiment to proceed without sedating drugs. In the absence of anesthesia, SWDs in GAERS were accompanied by increased thalamic and decreased cortical BOLD, matching the human studies.

After establishing conditions that reproduce, as much as possible, human absence seizures, the authors measured changes in single neuron firing in 2 brain regions previously shown to be critically engaged in rodent absence seizures, the subgranular somatosensory cortex and thalamic ventrobasal (VB) nucleus (ventroposterior medial and ventroposterior lateral nuclei). When the activity of all neurons was measured together, both cortical and VB thalamic neurons showed a decrease in mean firing rates during the SWD with an increase in firing rate during the spike component but a greater decreased rate during the wave component. Importantly, these measurements revealed 4 distinct types of cortical and thalamic neuronal firing patterns in relation to SWD onset (1) approximately 42% of neurons had reduced firing rate throughout the whole SWD (sustained decreased), (2) approximately 6% exhibited an increased firing rate throughout the discharge (sustained increase), (3) approximately 22% had increased firing at SWD onset, but baseline firing throughout the remainder of the discharge (onset peak), and (4) the remaining neurons had no change in firing rates. Remarkably, the 4 different neuron types were exceptionally stable and exhibited the same firing pattern with all recorded seizures (∼50-200 recorded SWDs).

Prior EEG studies of mouse and rat SWDs found that absence seizures were not instantaneous events beginning with the first EEG spike but were associated with EEG changes that began seconds prior to the first SWD spike. ^{6 -8} McCafferty et al found decreased high frequency (>40 Hz) and increased low frequency (<40 Hz) EEG power beginning much earlier, 40 to 60 seconds prior to the SWD. These pre-ictal changes in EEG power were accompanied by decreased neuron firing and licking rates.

This paper is noteworthy for the novel experimental designs used to test behavior and fMRI aspects of rodent absence seizures as well as its finding of the 4 stable functional classes of cortical and thalamic neuronal firing during SWDs. The considerable effort spent training GAERS to perform the tasks in a setting of relaxed wakefulness and to tolerate fMRI studies without anesthesia was crucial for obtaining results that can directly link GAERS SWDs to human absence seizures. Undoubtedly, these behavioral paradigms will be utilized in future studies of animal absence epilepsy.

The result that GAERS SWDs confer impaired awareness strengthens our confidence that we can continue to use these rodent models to develop new therapies for generalized seizures. If these discharges did not truly depress consciousness, there would be justified concern that they did not engage the same networks as human seizures and thus therapies effective in rodents may not translate to patients. Likewise, the finding that unanesthetized GAERS SWDs increase thalamic- and decrease cortical BOLD signal reassures investigators that the rodent SWDs are acting via a similar mechanism as patients.

The authors’ finding of decreased neuronal activity during SWDs corresponds to the reduced cortical fMRI BOLD signal. However, it is surprising that thalamic BOLD increased while neuronal firing decreased. The authors suggest that this increased thalamic neurovascular coupling may originate from massive postsynaptic inhibitory potentials in VB. Although it was not discussed in the paper, it is also possible that the BOLD signal originates from a nearby thalamic nucleus. For example, a previous study found increased SWD-associated neuronal firing in the nearby thalamic reticular nucleus. ⁹ Future studies will ultimately determine the source of increased thalamic BOLD signal but this apparent discrepancy in fMRI/electrophysiology findings highlights the necessity of continuing nonhuman studies with invasive single unit electrophysiology rather than relying solely on human fMRI experiments and inferring the physiology.

Previous studies identified reduced mean neuronal firing rates during SWDs in GAERS ⁹ and stargazer mice. ¹⁰ However, a notable finding of this paper was the observation of 4 distinct and stable patterns of neuronal firing in both cortex and VB. Why, in a single thalamic nucleus and within the subgranular layers of a small region of somatosensory cortex, do neurons have such divergent responses in SWDs? One obvious answer would be that the 4 different functional classes correspond to different neuronal types. For example, cortical fast spiking inhibitory neurons could constitute the sustained increased firing class. Unfortunately, the authors could not identify the recorded neurons with histology or separate excitatory and inhibitory units with spike sorting and thus the identification of the neurons within the 4 functional classes needs to be performed in future research. Nevertheless, given that interneurons only comprise 15% to 20% of neurons in the cortex and may not even be present in VB, it is unlikely that the 4 neuron classes will be attributed solely to excitatory neurons and different types of interneurons. Rather, it is more likely that differences in connectivity or intrinsic biophysical properties contribute to the functional diversity. Regardless of their origin, the existence of 4 physiologically distinct types of neurons suggests that they could be selective targets of pharmacological and/or neurostimulation therapies. Moreover, the finding that the change in single unit firing rates may occur up to a minute before seizure onset could make a responsive neurostimulation option especially promising.

In conclusion, this paper is significant for identifying SWD-associated changes in fMRI BOLD responses, demonstrating SWD-associated altered awareness and identifying 4 distinct and stable functional classes of neurons. These findings will advance the discovery of new pharmacological and network stimulation treatments in rodent models of generalized epilepsy for translation to human patients.

Martin J. Gallagher MD, PhD
Department of Neurology, Vanderbilt University School of Medicine