Serotonergic Dysregulation in Sudden Unexpected Death in Epilepsy
Frida A. Teran, Rup K. Sainju, Eduardo Bravo, Jacy Wagnon, YuJaung Kim, et al.
(see pages 4959–4971)
Sudden unexpected death in epilepsy (SUDEP) is rare, affecting about 1 in 1000 people with epilepsy, but it may be preventable. SUDEP has been linked to respiratory failure, but the cellular mechanisms behind it remain unknown. SUDEP often follows a generalized convulsive seizure (GCS), with death caused by postictal respiratory arrest. This week, Teran et al. show that the problem lies with a population of serotonergic neurons tasked with chemoreception and thermoregulation. Respiration is driven by wakefulness signals from the forebrain and by chemoreceptors that monitor CO2, both of which are susceptible to disruption by GCS. The authors had previously established the hypercapnic ventilatory response (HCVR) as a measure of respiratory carbon dioxide (CO2) chemoreception. Here, they measured interictal HCVR in 83 epileptic patients, 7 of whom went on to experience a seizure while undergoing monitoring. In eight of nine seizures, the HCVR was lower postictal versus interictal, indicating a reduced capacity to sense CO2. Wild-type mice and mice with Dravet Syndrome (DS), an animal model of refractory epilepsy, displayed similar HCVR at baseline, but following a heat-induced seizure in the DS mice, the HCVR was decreased significantly. The authors reasoned that if chemoreception had been affected by a dysfunction in serotonergic neurons, thermoregulation would also be affected. Indeed, mice that displayed a large decrease in HCVR also showed a larger drop in body temperature postictally. Depletion of serotonin with the tryptophan hydroxylase inhibitor PCPA (4-chloro-dl-phenylalanine methyl ester hydrochloride) produced similar defects in CO2 chemoreception and thermoregulation to that seen postictally in DS mice and increased mortality following heat-induced seizures. Conversely, the serotonin agonist fenfluramine administered 45 min before seizure induction rescued the impairments seen in postictal chemoreception and thermoregulation. Together the findings elucidate the cellular underpinnings of respiratory failure that causes SUDEP and may point toward future serotonergic prophylactic therapies.
Perceptual model representational dissimilarity matrices (mRDMs) for auditory and two VT stimuli representing 9 trained words.
Evidence for Metamodal Encoding in Neurotypical Adults
Srikanth R. Damera, Patrick S. Malone, Benson W. Stevens, Richard Klein, Silvio P. Eberhardt, et al.
(see pages 4984–4996)
Neuroscience dogma holds that the brain is organized according to its sensory modality inputs, so the auditory cortex is attuned specifically to sound, and so on. But a newer thesis holds that the brain is organized in a more task-centric way, putting to use whatever sensory stimuli come along. In support of this so-called “metamodal,” sensory-independent organization are studies of people who have lost a sense, such as blind people who use echolocation to navigate in the world. Fewer studies, however, have investigated the possibility of metamodal brain organization in people who are fully sensorily abled, and the results have been mixed, so the thesis has not been accepted as a general principle of brain organization. Now Damera et al. demonstrate evidence for metamodal organization in neurotypical adults. The authors used a clever experimental design in which participants were trained to recognize words encoded as vibrotactile (VT) stimuli to the forearm. Participants received training with either a token-based algorithm or a vocoded (or voice-encoded) algorithm, in which characteristics of the words were retained. Functional magnetic resonance imaging during auditory stimulation showed activation of the superior temporal gyrus (STG), an area known to encode auditory speech, whereas VT stimuli evoked activity in the supplementary motor area and precentral gyri. Remarkably, the trained VT stimuli based on vocoded speech were encoded in the mid-STG similar to auditory speech, but token-based VT stimuli were not. In participants who received the vocoded training, but not the token-based training, functional connectivity was also increased between somatosensory and auditory speech areas. Participants receiving either type of training performed equally well in recognizing the VT-based words; the researchers wondered what circuitry did underlie the token-based learning. Those who received token-based training showed an effect of training on hippocampal connectivity, likely reflecting the neural correlates of associative learning. The results showed that VT stimuli could be metamodally encoded in auditory brain areas, but that metamodal engagement in neurotypical individuals requires that the neural encoding of stimuli be compatible between the established and the novel sensory modality. These findings may contribute to the development of sensory substitution devices for individuals who are unable to detect certain stimuli.
Footnotes
This Week in The Journal was written by Stephani Sutherland
