Using Magnetic Resonance Spectroscopy in Narcolepsy to Study the Limbic Mechanisms of Cataplexy

THERE ARE SEVERAL REASONS TO SUSPECT THAT LIMBIC STRUCTURES IN THE BRAIN SUCH AS THE AMYGDALA ARE INVOLVED IN THE PATHOPHYSIOLOGY of narcolepsy, and especially in the generation of cataplexy. If cataplexy is an intrusion of the normal muscle atonia present during rapid eye movement (REM) sleep into wakefulness, one of these reasons is the involvement of the same limbic structures in the generation of REM sleep. In this issue of Sleep, Poryazova et al.¹ take a critical step in the study of human narcolepsy by examining the amygdala in an a priori manner using magnetic resonance spectroscopy.

Narcolepsy (with cataplexy) is a sleep disorder that affects approximately 100,000 to 250,000 Americans. The primary symptoms of narcolepsy are excessive daytime sleepiness and cataplexy. Narcolepsy is a neurodegenerative disease that is caused by damage to hypocretin neurons in the hypothalamus. The cause of this damage can be purely genetic,² purely environmental,³ or—much more commonly—some gradation between these two extremes.⁴

Sleepiness is the most debilitating symptom of narcolepsy, but cataplexy is the most unusual. Many people mistake cataplexy for a sudden attack of sleep. Instead, it is a loss of voluntary muscle tone triggered by emotions without a loss of consciousness. Cataplexy is typically viewed as an intrusion of the normal muscle atonia present during REM sleep into wakefulness. Another theory relates cataplexy to tonic immobility,⁵ though these theories may be compatible since REM sleep and tonic immobility have similar neural mechanisms.

The emotional triggers of cataplexy can be understood in terms of the emotional nature of dreaming. While some form of mentation can occur throughout sleep, prototypical dreaming occurs in REM sleep.⁶ Dreams arising from this stage are highly emotional.⁷ This is consistent with the increased activity in limbic (i.e., related to emotion) brain regions during human REM sleep.^8–10 One limbic region of particular interest is the amygdala, which may be involved in both cataplexy¹¹ and REM sleep.¹²

Poryazova and colleagues tied these ends together by studying the integrity of the amygdala—as well as the hypothalamus and pontomesencephalic junction—in narcolepsy patients using magnetic resonance spectroscopy. This technique is based on the same principles as magnetic resonance imaging. Instead of examining the absorption and emission of electromagnetic energy in gray versus white matter, this technique examines a single region and creates an energy spectrum where the peaks correspond to the unique resonance frequency of each chemical in that region.

Poryazova et al. focused their efforts on two chemicals: N-acetylaspartate and myo-Inositol. They discovered a decrease of myo-Inositol in the right amygdala of narcolepsy patients compared to controls. However, they hypothesized that there would be an increase in this chemical since it is a marker of gliosis. Given it is also involved in intracellular signaling, Poryazova et al. suggested that this decrease may reflect an alteration of this process.

An advantage of magnetic resonance spectroscopy and neuroimaging techniques in general is you can assess multiple regions of interest as well as the interaction between those regions. Poryazova et al. correlated the concentration of the chemicals in one region with the concentration of the chemicals in another region. As the levels of N-acetylaspartate in the hypothalamus decreased, the levels of myo-Inositol in the right amygdala increased. This correlation was only present in the patient group.

This may mean that as the integrity of hypocretin neurons in the hypothalamus decreases, the downstream effects of this degeneration (via the missing hypothalamo-amygdala projections) in the amygdala increases. Such an abnormal interaction between these regions could be a mechanism of cataplexy if it is triggered by the amygdala. This idea fits with the results of a functional magnetic resonance imaging study of humor processing in narcolepsy patients by the same group¹³: activity in the hypothalamus was decreased relative to controls while activity in the amygdala was increased relative to controls.

Poryazova et al.¹ should be highly commended for studying the pathophysiology of human narcolepsy by examining brain regions that are known to play a role in emotional processing and in REM sleep, as others have done in canine narcolepsy.¹¹ This study also represents the standard that all human narcolepsy studies should attempt to match in terms of patient selection and description.

While the negative hypothalamo-amygdala correlation is a very interesting finding, some normal amount of caution is warranted. The authors' interpretation depends on the idea that low levels of N-acetylaspartate are “bad” (since it is a marker of neural integrity), whereas high levels of myo-Inositol are “bad.” However, myo-Inositol can be viewed as either “good” or “bad” depending on whether it reflects normal intracellular signaling or gliosis and other correlations were found that might further confuse the interpretation of the results as a whole.

Poryazova et al.¹ used an approach that results in a relative quantification of these chemicals as they were all expressed relative to creatine. This is normally a safe assumption because creatine has the most stable concentration. Future research might refine this approach by obtaining absolute concentrations since creatine levels could be altered in a neurodegenerative disease like narcolepsy.

Returning to this group's publication on humor processing in narcolepsy patients, in addition to the increased activity in the amygdala relative to controls, there was also increased activity in the nucleus accumbens relative to controls. This area is the destination for neurons in the mesolimbic dopaminergic pathway. Although the importance of dopamine as a sleep regulator was diminished in the past due to the apparent stability of the firing rate of these neurons across the sleep-wake cycle, this sentiment is no longer tenable.¹⁴ Dopamine levels in the nucleus accumbens—as measured by intracerebral microdialysis—decrease during slow-wave sleep and return to waking levels during REM sleep.¹⁵ This region may be another limbic region worthy of study in human narcolepsy.¹³

While emotions do not trigger REM sleep per se, dreams do trigger emotions, and perhaps the associated limbic structures trigger the muscle atonia of REM sleep. If emotions affect the same motor neurons that are involved in cataplexy, then perhaps this is why emotions trigger cataplexy.

DISCLOSURE STATEMENT

Dr. Picchioni has indicated no financial conflicts of interest.