Pushing and pulling with no end in sight! The role of cross-talk between different forms of respiratory plasticity in modifying sleep apnea.

Therapeutic exposure to intermittent hypoxia leads to long-term facilitation (LTF) of diaphragm and upper airway muscle activity, resulting in sustained increases in minute ventilation and improved upper airway stability (Mateika and Narwani 2009). These findings led to the hypothesis that exposure to intermittent hypoxia, which accompanies hypopneic or apneic events during sleep, could initiate LTF of chest wall (e.g. diaphragm and inspiratory intercostal muscles) and upper airway muscle activity. Initiation of LTF could serve to stabilize breathing and the upper airway, leading to reductions in breathing events across the night or life-span (Mateika and Narwani 2009). However, there is little support for this hypothesis, since many studies conducted over decades have reported that the severity of sleep apnea increases from the evening to the early morning. The prevailing hypothesis is that LTF does not typically manifest itself in individuals with sleep apnea because of the withdrawal of chemoreceptor input (Harris, Balasubramaniam et al. 2006). This withdrawal is induced by the presence of hypocapnia that occurs in response to hyperventilation that is elicited when emerging from apneic events. However, there are other possibilities that could explain the ineffectiveness of respiratory plasticity in mitigating apneic events during sleep.

Along these lines, Fields and colleagues (2019) explored if initiation of competing forms of plasticity, induced by repetitive apnea, prevent the induction of sustained increases in phrenic nerve activity. In addition to LTF induced by intermittent hypoxia, recurrent reductions in respiratory neural activity initiates another form of plasticity, known as inactivity-induced inspiratory motor facilitation (iMF) (Streeter and Baker-Herman 2014). iMF is robustly expressed in phrenic, inspiratory intercostal and hypoglossal motor pools. Given that central sleep apnea is characterized by periods of reduced or inactive motor activity, and mixed apneic events are characterized by motor inactivity at the onset of an event, the authors proposed that iMF may be initiated during sleep (Fields, Braegelmann et al. 2019). Given that hypoxia typically accompanies motor inactivity, two forms of respiratory plasticity could theoretically be initiated in response to apneic events. However, there is little evidence in humans that compensatory breathing adaptations clearly manifest (Mateika and Narwani 2009). Thus, Fields and colleagues (Fields, Braegelmann et al. 2019) hypothesized that the simultaneous initiation of LTF and iMF prevents the expression of compensatory modifications.

To test the hypothesis, a series of clever experiments using anesthetized, paralyzed, vagotomized and ventilated rats were employed (Fields, Braegelmann et al. 2019). In one set of experiments, LTF of phrenic nerve activity was initiated by intermittent hypoxia, which was induced by ceasing mechanical ventilation 5 times for 25 s. In another series of experiments, iMF was initiated by eliminating motor activity by decreasing carbon dioxide levels 5 times for 25 s. Thereafter, the two perturbations were combined (i.e. carbon dioxide was reduced in the absence of mechanical ventilation). Although the separate stimuli initiated the two forms of plasticity as expected, there was no evidence of compensatory mechanisms when the two stimuli were combined (Fields, Braegelmann et al. 2019). This observation is exciting and is tied in with our previous hypothesis, which suggested that increases in carbon dioxide levels are required for the manifestation of LTF (Harris, Balasubramaniam et al. 2006). In other words, if hypocapnia reduces or eliminates respiratory motor activity during sleep, then iMF may be initiated concurrently with LTF leading to occlusion of compensatory mechanisms that mitigate apnea severity (Fields, Braegelmann et al. 2019).

Fields and colleagues (Fields, Braegelmann et al. 2019) went a step further to explore the mechanistic underpinning of the interaction between LTF and iMF. Based on previous work (Streeter and Baker-Herman 2014), they hypothesized that NMDA receptor activation could be responsible for the absence of compensatory breathing adaptations in individuals with sleep apnea. This hypothesis was based on the understanding that spinal NMDA receptor activation is necessary for the induction and maintenance of LTF, but constrains iMF expression (Streeter and Baker-Herman 2014). This possibility was supported by their findings which showed that administration of an NMDA receptor antagonist prevented the initiation of LTF and lead to sustained increases in phrenic activity when intermittent hypoxia and respiratory motor inactivity were combined (Fields, Braegelmann et al. 2019). Given these findings, the application of NMDA receptor antagonists could serve as a pharmacological intervention for sleep apnea.

However, Fields and colleagues (Fields, Braegelmann et al. 2019) recognized that the use of NMDA antagonists may not be useful therapeutically because of accompanying psychotropic side effects. Thus, the role of retinoic acid in promoting the initiation of compensatory responses following exposure to intermittent hypoxia and respiratory motor inactivity was explored (Fields, Braegelmann et al. 2019). The inhibition of retinoic acid synthesis prevented iMF but had no effect on LTF (Fields, Braegelmann et al. 2019). They also showed that the application of retinoic acid resulted in the manifestation of compensatory responses following intermittent motor inactivity combined with intermittent hypoxia (Fields, Braegelmann et al. 2019). Thus, retinoic acid could serve as therapeutic intervention to allow for the manifestation of compensatory forms of plasticity that might ultimately contribute to reducing the severity of sleep apnea.

Overall, Fields and colleagues (Fields, Braegelmann et al. 2019) have set the stage to explore if iMF is initiated in humans with sleep apnea, which would complement our work that has shown that LTF can be initiated in men and women during wakefulness and sleep under controlled laboratory conditions. The challenge will be to demonstrate its existence under natural conditions in the presence of other mitigating factors (i.e. differences in sleep stages, sleep fragmentation in response to arousal), that are likely to influence the manifestation of iMF. If iMF is an observed phenomenon in humans, the next step will be to explore the interaction between LTF and iMF. In our opinion, the exploration of these possibilities will likely manifest most clearly in individuals with central sleep apnea. Indeed, a simple but significant beginning would be to explore if iMF is present in individuals with central sleep apnea that are administered supplemental oxygen to prevent hypoxemia.