Seizing upon mechanisms for impaired consciousness

Summary

Transient loss of consciousness experienced during, and shortly after, focal temporal lobe seizures is a complex phenomenon with life threatening repercussions for persons with epilepsy. In this issue of Neuron, Motelow et al. (2015) describe the first evidence for decreased cholinergic drive and depressed subcortical arousal in seizures as a novel mechanism for impaired cortical function.

The controversial subject of consciousness dates back at least as far as the philosophical theories of Descartes in the seventeenth-century. However, recent advances in neuroscience techniques to measure neural activity in both human and animal models have begun to shed new light on this ancient affair. Consciousness comprises the state of wakefulness, awareness, or alertness in which sentient beings function while not asleep. Although the exact mechanisms that mediate consciousness are poorly understood, studies correlating brain activity with states of consciousness have identified widespread bilateral cortical networks and subcortical arousal systems as key players. In temporal lobe epilepsy (TLE), complex-partial seizures are those associated with impaired consciousness both during, and after, the seizure, unlike simple-partial seizures, which do not disrupt consciousness (Penry, 1981). Investigators have long been bewildered as to how a focal seizure confined to the temporal lobe, a brain region not typically thought of as essential for consciousness, can nevertheless profoundly impair consciousness. Clearly other structures are involved.

Impaired consciousness, such as during deep sleep or coma, typically produce cortical slow wave activity, decreased cerebral blood flow, and reduced metabolism (Steriade et al., 1993; Cowan and Wilson, 1994; Haider et al., 2006). Similarly, impaired consciousness in complex-partial seizures is highly correlated with ictal and postictal slow wave activity and abnormal increased cerebral blood flow in the upper frontal and parietal neocortices (Lieb et al., 1991; Blumenfeld et al., 2004a; Blumenfeld et al., 2004b), as wells as brainstem and medial diencephalon (Lee et al., 2002; Blumenfeld et al., 2004b; Tae et al., 2005). Consistent with the results of electrophysiological and neuroimaging studies in human TLE, impaired consciousness (as measured by behavioural arrest) in rat during partial limbic seizures is associated with slow wave activity, reduced neural activity, and decreased metabolism in the frontal cortex (Englot et al., 2008). Animal models of TLE have allowed detailed mechanistic study of the role of subcortical structures in impaired consciousness. For example, stimulation of the septum results in the appearance of cortical slow waves and behavioural arrest, whereas, fornix lesions blocked seizures from reaching subcortical structures such as the septum and thalamus (Englot et al., 2009) and prevented cortical slow waves and the associated behavioural changes during seizures. Englot et al. concluded that activity in subcortical structures, in particular the septum, is required for ictal neocortical slow activity and behavioural arrest in partial seizures. Together, these previous animal studies have documented a remarkable correlation between reduced neural activity in the frontal cortex and subcortical structures and the impaired consciousness during complex partial seizures. However, the specific circuit mechanisms underlying impaired consciousness in complex partial epilepsy have been elusive, with the answer to the following question still unresolved: how might seizure activity in the temporal lobe lead to activation of these subcortical structures and result in neocortical dysfunction and impaired consciousness?

Various hypotheses for loss of consciousness during complex partial siezures have been proposed. One suggests that seizure-related impairment results from spread of ictal activity to bilateral temporal lobes (Inoue and Mihara, 1998). An alternate view is the “network inhibition hypothesis”, which suggests that seizures spread from the temporal lobe to activate inhibitory subcortical structures that in turn deactivate frontal cortical regions necessary for the normal conscious state (Norden and Blumenfeld, 2002). However, this hypothesis relies on an unidentified major inhibitory component(s) within the poorly understood consciousness network and is based in part on a correlative, not causative, link between slow wave activity and impaired consciousness.

In this issue of Neuron, Motelow et al. bridge a gap in the network inhibition hypothesis and uncover novel mechanisms by which subcortical arousal circuits are depressed during focal seizures associated with impaired consciousness. Along with increased activity in the hippocampus, anterior hypothalamus, and lateral septum, the authors report depressed activity in the frontal cortex as well as subcortical arousal structures, including intralaminar thalamus and midbrain tegmentum. It is well established that withdrawal of acetylcholine signalling plays a critical role in cortical slow wave activity observed during sleep and in depression of thalamic and cortical activity (Marrosu et al., 1995). Motelow et al. perform in vivo juxtacellular recordings and demonstrate a decrease in firing of all recorded cholinergic neurons within the pedunculopontine tegmental (PPT) nuclei and the basal forebrain, nuclei that indirectly (via thalamus) and directly activate the cortex, respectively, during seizure activity. Importantly, in contrast to cholinergic neurons, noncholinergic neurons in both regions displayed mixed changes in firing rates during seizures. A decrease in extracellular choline concentration measured in both the cortex and thalamus, targets of cholinergic projections from PPT and basal forebrain, was also detected during limbic seizures. These results provide the first and only solid evidence that cholinergic transmission from the subcortical arousal structures to the thalamus and cortex is impaired during seizures.

Furthermore, these results motivate new and exciting questions as to the role of the depressed cholinergic inputs to thalamus and cortex during seizures with impaired consciousness. For example, is the reduced cholinergic activity required and sufficient for the seizure-induced impaired consciousness? If so, can cholinergic activity be targeted to prevent impaired consciousness? In a very recent study, Glummadavelli et al. reduced cortical slowing and promoted behavioural arousal by electrically stimulating the downstream cholinergic targets in the intralaminar thalamus (Glummadavelli et al., 2014). Could stimulation of the PPT also reduce cortical slow wave activity and enhance arousal?

The present study makes a critical contribution to the current network inhibition hypothesis. However, there are several key components missing in this hypothetical model. The neural basis for the “reverse polarity switch”, responsible for converting the seizure-related activation of brain structures (ie. hippocampus, anterior hypothalamus, and lateral septum) into suppression of subcortical arousal structures (ie.PPT, intralaminar thalamus, and basal forebrain) and, ultimately, frontal cortex, remains unknown (Fig. 1). Characterizing the circuit mechanism(s) responsible for the suppressed cholinergic neurons in the PPT and the basal forebrain is one possible route to unveil the identity of the reverse polarity switch (Fig. 1)? The lateral septal nuclei display increased neural activity during seizures and consist mainly of inhibitory neurons that send widespread projections to a plethora of subcortical structures making it an attractive candidate for suppression of the subcortical arousal system. Although the septum has been shown to be important for neocortical slow waves and arousal during seizures, the direct downstream targets important for these activities have not been identified. Lastly, the subcortical arousal system is composed of several components, which begs the question: What other structural/neurochemical components are modulated during seizure activity and what is their role in cortical dysfunction and consciousness?

Figure 1.

Adaptation of “Network Inhibition Hypothesis”

Seizure activity spreads from the hippocampus to activate lateral septal nuclei and anterior hypothalamus. Putative efferents from the aforementioned structure(s) activate an unknown reverse polarity switch, converting signal propagations from excitatory to inhibitory. Activation of the unknown reverse polarity switch results in reduced cholinergic drive from subcortical arousal structures (PPT) to intralaminar thalamus and basal forebrain to frontal cortex. The reduced cholinergic transmission directly and indirectly triggers slow wave activity in the frontal cortex and impairs consciousness. What remains unknown are the roles of other subcortical arousal systems likely to contribute to this network.

The discovery that cholinergic depression in subcortical arousal structures is correlated with focal seizures paves the way for future experiments to dissect the processes involved in impaired consciousness. As with many studies of experimental epilepsy, future studies that elucidate mechanisms involved in cortical dysfunction and epilepsy-related impairment of consciousness will undoubtedly give insight into the basic mechanisms and circuitry involved in normal brain activity, such as consciousness itself. Although seizure prevention is the primary goal of epilepsy research, therapeutic intervention aimed to prevent comorbidities, including consciousness impairment, is a critically important goal for enhancing the quality of life for those living with epilepsy.