Mechanisms of Upper Airway Hypotonia During Sleep

In healthy humans and animals with non-collapsible upper airway, upper airway muscles have a low level of activity during quiet wakefulness, minimal or no activity during non-REM sleep, and very irregular, often high but transient, bursts of activity during REM sleep. This variability may be related to the overall highly phasic brain activity during REM sleep and to the presence of dreams which occur this stage of sleep.

Kubin Figure 5When present, activity of many upper airway muscles and motoneurons is respiratory modulated. The rhythmic activation coincident with the inspiratory phase of the respiratory cycle stiffens and widens the airway at the time when the negative inspiratory pressure inside the airway exerts a collapsing force. Nevertheless, steady (tonic) inputs to upper airway motoneurons also importantly protect the airway wall from collapse. Overall, the level of motoneuronal activity at any given time and behavioral state is determined by the balance of excitatory and inhibitory inputs that originate in three functionally distinct sources: central respiratory, central tonic, and reflexive (from peripheral mechano- and chemoreceptors). We and others have determined that the activity of those central neurons that provide inspiratory activation to hypoglossal (XII) motoneurons is not reduced during REM sleep while activity of XII motoneurons is depressed. This suggested, and our studies have confirmed, that the magnitude of tonic, rather than inspiratory, excitatory inputs to motoneurons is reduced during sleep, and REM sleep in particular.

Kubin Figure 6The example here shows a semi-continuous tracing of the magnitude of the tongue (lingual) and postural muscle (nuchal) electromyograms (EMGs) in a chronically instrumented rat (hence, an animal with non-collapsible upper airway) in relation to changes in behavioral states, as determined in successive 10 s intervals.  During the periods of wakefulness (W; grey symbols), lingual EMG is variable, with the highest activity corresponding to the periods of active behaviors. During slow-wave (non-REM) sleep (SWS; blue symbols), lingual activity is almost entirely abolished, whereas following each transition into REM sleep (REMS; red symbols) lingual EMG increases greatly, with the increase caused by frequent and large bursts of activity.  See: Lu et al. (2005).

In contrast to subjects with non-collapsible airway, upper airway muscle tone in OSA patients is greatly elevated even during quiet wakefulness. This compensatory increase helps maintain the airway open and is partially carried over into non-REM sleep. This is one reason why the changes in upper airway motor tone across behavioral states occur with a different pattern in normal subjects and in OSA. To identify the mechanisms that contribute to sleep-related suppression of upper airway motor tone in OSA patients, we use anesthetized rats in which upper airway muscle tone is acutely elevated by elimination of certain peripheral inhibitory reflexes (vagotomy). In combination with the use of an acute pharmacological model of REM sleep, this allows us to explore the mechanisms of the depressant effects of REM sleep on upper airway muscle activity.

We have identified some of the most important non-respiratory excitatory inputs to upper airway motoneurons whose magnitude is gradually reduced and ultimately eliminated during transitions into non-REM and then REM sleep. Specifically, the amount of serotonin (5-HT) released onto XII motoneurons is be reduced during sleep because the 5-HT-containing medullary cells that send axons to the XII motor nucleus have reduced activity during non-REM sleep and become silent during REM sleep (Kubin et al., Brain Res. 1994, 645: 291-302; Woch et al., J. Physiol. (Lond.) 490: 745-758, 1996). Norepinephrine (NE)-containing cells have the same pattern of changes in activity across the states of wakefulness and sleep, and they are also silenced during REM sleep-like state elicited pharmacologically in anesthetized rats. See Carbachol models of REM sleep. As is the case with 5-HT, noradrenergic fibers make synaptic contacts with XII motoneurons, and NE excites these motoneurons.

Kubin Figure 7Thus, a combined loss of serotonergic and noradrenergic activation may be a major cause of the depression of upper airway muscle activity during sleep, and especially REM sleep. The process involving a withdrawal of excitatory effects that results in diminished or eliminated output activity is called disfacilitation. Our studies in the anesthetized rat model of REM sleep indicate that the sleep-related loss of activity in upper airway muscles can be fully explained as resulting from a combined withdrawal of noradrenergic and serotonergic excitation (Fenik et al., Am. J. Resp. Crit. Care Med. 172:1322-1330, 2005).

The positive evidence for a major role of a combined noradrenergic and serotonergic disfacilitation contrasts with multifaceted evidence from our lab and other groups that glycinergic and GABAergic inhibitions do not contribute in any significant way to the depression of upper airway muscle tone during REM sleep (Kubin, L. 2008. Sleep 31:1473-1476, 2008).