Breathing Disturbances in Rett Syndrome

Abstract

Rett syndrome is an X-linked neurological disorder characterized by behavioral and neurological regression, seizures, motor deficits and dysautonomia. A particularly prominent presentation includes breathing abnormalities characterized by breathing irregularities, hyperventilation, repetitive breathholding during wakefulness, obstructive and central apneas during sleep, and abnormal responses to hypoxia and hypercapnia. The condition and pathology of the respiratory system is further complicated by dysfunctions of breathing-motor coordination, which is reflected in dysphagia. The discovery of the X-linked mutations in the MECP2 gene has transformed our understanding of the cellular and molecular mechanisms that are at the root of various clinical phenotypes. However, the genotype-phenotype relationship is complicated by various factors which includes not only X-inactivation, but also consequences of the intermittent hypoxia and oxidative stress associated with the breathing abnormalities.

Introduction

Rett Syndrome is a neurological disorder characterized by severe breathing abnormalities as well as a series of behavioral regressions that become apparent at the age of 6 months (1, 2). The behavioral regressions lead to a slowing of head circumference growth, the loss of language, gait and purposeful hand use, EEG abnormalities, seizures, spasticity, scoliosis, and the children typically develop hand stereotypies such as hand wringing or hand clapping (3, 4). Many of these key clinical features were first described in 22 girls by Andreas Rett in 1966 (5–7). Rett syndrome is the second most prevalent genetic cause of intellectual disabilities in girls. Most cases are caused by mutations in the MECP2 gene (8). This gene is located on the X-chromosome and is critical for DNA methylation, and functions as a transcriptional regulator of many other genes. X-inactivation (9) (10) and the fact that MECP2 regulates many other genes, make this disorder and its clinical phenotype highly complex. Adding to the complexity is the fact that MECP2 mutations are neither necessary nor sufficient for diagnosing Rett Syndrome, thus Rett syndrome remains a clinical diagnosis (11, 12). Variants of atypical Rett Syndrome have been described by mutations in different loci, including CDKL5 or FOXG1 mutations (13, 14). Rett Syndrome is primarily observed in females, yet a small number of males with MECP2 mutations have been described and characterized as Male Rett Encephalopathy. The clinical phenotype of this encephalopathy is severe and boys typically die by the age of 2 (15). A natural history study has provided deep insights into the Rett Syndrome classification and the various clinical phenotypes (4, 16–21). Unfortunately, despite major scientific breakthroughs, most clinical trials have failed, and the currently available therapies offer only very limited success (6, 20, 22–29). Thus, the devastating breathing abnormalities, like most other symptoms, are largely resistant to pharmacological treatment.

However, promising novel therapeutic approaches (30–32), and research in animal models raise considerable hope. Animal models exhibit many of the clinical symptoms and various treatments targeting neurons and glia have successfully reversed breathing disturbances and increased the lifespan of mutant mice (33–48). This book chapter will focus primarily on the breathing abnormalities in Rett Syndrome. Like many of the other clinical phenotypes, the breathing dysfunctions are complex, largely resistant to standardized care and influence autonomic functions (dysautonomia)(49–56). Dysautonomia is the consequence of interrelated disturbances in cardiorespiratory coupling and other vital autonomic functions that may be a major contributor to the sudden death as has been reported for Rett Syndrome (57–64). Studies in mouse models suggest that dysautonomia depends on the expression of MECP2 in cholinergic neurons. The expression in these neurons is necessary and sufficient for autonomic cardiac control, thermoregulation, and survival (64). However, it is important to emphasize that additional factors and disturbances contribute to a decreased life span and dysautonomia (65, 66).

Irregular Breathing and Breathholds in Rett Syndrome

Breathing abnormalities are experienced by more than 80% of Rett Syndrome patients during their lifespan (17, 67, 68). However, respiratory dysfunctions come and go. Most breathing problems occur at 6-11 years of age, often disappear between the ages of 11-13 years, and then relapse at any age (17). The reason for the appearance and disappearance remains unknown. Early studies recognized that the most obvious breathing dysfunctions occur during the day and a large catalogue of abnormalities have been described including hyperventilation, Valsalva maneuvers, apnea, apneusis, breathholding and rapid shallow breathing (52, 69–75). A breath-to-breath analysis reveals that both breathing frequency and tidal-volume are irregular. Breaths can be shallow, but sometimes also forceful. On average, breathing frequency, median airflow and heart rate are higher compared to healthy control children. However, average values do not adequately reflect the large individual variability and the fact that some children have normal breathing, and only occasionally exhibit breathing dysfunctions (49, 54). An increased variability in the breathing irregularities is also observed in the animal model for Rett Syndrome (76). In humans, one obvious source of this variability is the strong emotional component of breathing disturbances, as a multitude of symptoms can be interrelated and dynamically emerge. Repetitive rocking, screaming, agitation, anxiety, low mood, temperature abnormalities and gastrointestinal disturbances can be associated with the onset of breathing and cardiorespiratory disturbances (62). Indeed, this fragile emotional interrelationship is the main reason why the characterization of dysautonomia in Rett Syndrome is best conducted at home using non-invasive ambulatory monitoring technology which reduces stress and agitation induced by hospital settings (55). A frequently observed breathing disturbance in Rett Syndrome comes in form of breathholds (Fig. 1,2) (49). The breathholds begin with an inspiratory effort of normal duration which is followed by a brief postinspiratory phase during which heart rate decreases. Postinspiration is abruptly terminated by the onset of the actual breathhold phase and is concomitant with a dramatic increase in heart rate (Fig.2). This tachycardia is absent during breathholds of healthy subjects, that are characterized by a “diving-response” and bradycardia (77, 78). Thus, the breathholds in Rett patients are somewhat reminiscent of Valsalva maneuvers, i.e., forced exhalations against a closed glottis. Andreas Rett’s first descriptions of the breathholds use the German term “Pressatmung”, which refers to the pushing (“press”) of air against the glottis. In the breathholds of Rett Syndrome the heartrate increase persists for 2-3 more breaths following the resumption of normal breathing. The persistence of tachycardia is likely caused by various factors, including the increased chest pressure which triggers a series of events that include changes in the pulmonary circulation and cardiac output and a compensatory tachycardia (77, 79). While this reflexive component likely contributes to breathholds, the episodic nature of breathholds points to disturbances in CNS mechanisms. Separated by inter-breathhold intervals of varying duration, these breathholds cause intermittent oxygen desaturations, which may have similar detrimental consequences as the intermittent hypoxia caused by obstructive sleep apnea. The length of the inter-breathhold interval and the heart rate increase measured at exhalation seem to depend on the duration of the preceding breathhold (49). Interestingly, similarly looking episodic tachycardia events also occur during sleep in the absence of breathholds (Fig. 2), further supporting the notion that the heart rate changes are not caused by a simple Valsalva-like heart rate increase that is triggered by the chest pressure. Moreover, during nighttime the tachycardia precedes paradoxical breaths and does not follow breathholds, as seen in wakefulness. These considerations suggest that the episodic tachycardia occurring during the breathholds in wakefulness and spontaneously during sleep are the result of dysautonomia that originates within the CNS.

Fig. 1.

MECP2 s expressed largely throughout the brainstem, including areas within the pons and medulla involved in the neural control of breathing. In healthy cells, MECP2 locates to CPG islands to influence transcription regulation. Studies have shown deficiencies in MECP2 expression in patients presenting with Rett syndrome. This deficiency has been suggested to alter normal respiration and result in abnormal breathing patterns accompanied by overactive expirations, known as breatholding.

Fig. 2.

Examples of breathing patterns and breath holds in a representative child with Rett Syndrome during the day and night using inductance plethysmography. Respiratory irregularities and heart rate changes are observed during both the day and night. Sum amplitude (AMP), rib cage (RC), abdomen (AB), and heart rate (HR). Adapted from (49, 55).

Insights into the neuronal mechanisms underlying breathing dysfunctions were gained by studying Mecp2 knockout (KO) mice (80). Reminiscent of the clinical phenotype, these mice are characterized by respiratory irregularities and the episodic occurrence of prolonged respiratory cycles (56, 81, 82). These prolonged respiratory cycles and respiratory pauses occurring in wakefulness are often referred to as “apneas”. However, the use of this term is not intended as a mechanistic explanation for these respiratory pauses and should not imply that these pauses reflect central apneas, which are mechanistically defined by the cessation of respiratory drive and that are often seen in human Rett Syndrome during sleep (Fig. 3). The apneas occurring in wakefulness are likely the mouse-equivalent of the breathholds seen in humans which is associated with the activation of expiratory activity as described above. In the awake mutant mice the mean respiratory frequency is also altered, with some models showing an increase (36), and others showing a decrease (56).

Fig. 3.

A: A 30-second epoch from nocturnal polysomnographic recording. A representative central apnea event (highlighted in red) arising from NREM N3 sleep stage in a 6yr old girl with Rett syndrome. This event is characterized by a cessation of flow with associated lack of effort on the chest (CHEST channel) and abdomen (ABD channel) bands, followed by mild oxygen desaturation (highlighted in pink) and subsequent mild tachycardia.

B: A 30-second epoch from nocturnal polysomnographic recording. A representative obstructive apnea event (highlighted in orange), arising from NREM N3 sleep stage in a 6yr old girl with Rett syndrome. This event is characterized by cessation of flow associated with increased effort (CHEST channel) and abdomen (ABD channel) bands, followed by mild oxygen desaturation (highlighted in pink) and subsequent mild tachycardia. Abbreviations: NREM: Non-rapid Eye Movement; PFLOW: Pressure flow from thermistor; FLOW: Flow signal from thermistor; SUM FLOW: Summatory of PFLOW and FLOW channel signals; CHEST: Signal from chest band; ABD: Signal from abdominal band; OSAT: Oxygen saturation (%); PR: Pulse rate (BPM)

Critical for the breathing disturbances in Rett Syndrome, are the pontine Kölliker–Fuse nuclei (KF, (83)), the nucleus tractus solitarius (NTS, (37, 84)), locus coeruleus (85), and the ventral respiratory group (86). The KF is located in the dorsolateral pons and projects to the ventral respiratory column in the medulla. KF neurons regulate various aspects of breathing including postinspiratory activity and function as an inspiratory “off-switch” (87). Postinspiratory activity is particularly important for controlling laryngeal adductors and airflow modulation during speech, while glottal closure is needed to protect the lungs from noxious substances (83, 88) or aspiration during swallowing (89). The postinspiratory phase of vagus nerve activity initiates the breathhold and is followed by tonic discharge of hypoglossal nerve and abdominal activity generating the co-activation responsible for the air pressure that characterizes the breathhold (90, 91). The association of tonic hypoglossal nerve activity with active abdominal expiration and post-inspiratory vagal activity seen in the working-heart-brainstem in situ preparation (34) is reminiscent of the breathhold activities seen in patients. Disturbances in postinspiratory activity may also contribute to the loss of speech, and impaired swallowing as will be described in more depth later in this chapter (83, 92, 93). However, while the pons, and in particular the KF, is an important contributor to the breathholds, the medulla plays also a critical role in the origin of breathing dysfunctions as suggested by targeted genetic lesions (80).

Alteration in the chemosensory response

The pons, but also the medulla, plays a particularly important role in the chemosensitive response, which is altered in Rett Syndrome. In response to acute hypoxia, Mecp2 KO mice exhibit a sustained elevation of respiratory frequency which is different from the biphasic response of wild type mice that consists of an augmentation followed by a secondary depression (76, 94–96). The altered hypoxic response persisted if MECP2 function was selectively removed from inhibitory, excitatory and even catecholaminergic (tyrosine hydroxylase (TH)-containing) neurons located in the caudal pons and medulla defined by the HOXB1 gene (76). A normal hypoxic response was restored by re-expressing MeCP2 specifically in these neurons, which indicates that the altered hypoxic response originates in these brainstem regions (76). A likely contributor to the disturbed hypoxic response is the NTS (76), region that is critical for processing chemosensitive afferent information from the carotid body. It is thought that MECP2 likely contributes to an imbalance in excitatory and inhibitory processing in the NTS, known to be critical for the hypoxic response (97). The NTS-mediated afferent input regulates both medullary and pontine activity (98), and the consequences of a MECP2 mutation may be even more far-reaching, given that the NTS receives important peripheral sensory inputs, not only from the carotid body, but also from lung mechanoreceptors (37, 56, 99–102). Interestingly, restoring MECP2 in the brainstem regions defined by the HOXB1 gene rescued only the hypoxic response, but not basal hyperventilation. To rescue basal ventilation, it was necessary to restore MECP2 in structures defined by the HOXA4 domain. These brainstem-selective lesion and rescue experiments support the notion that the maintenance of breathing requires a distributed respiratory network, which includes not only the medulla, but also the pontine respiratory group as discussed above (80).

Mutant mice also exhibit a reduced ventilatory CO₂ sensitivity (82, 95), which is consistent with disturbances in CO₂ homeostasis in human patients (103). The reduced sensitivity seems to involve astrocytes. Astrocytes express MECP2 (104, 105), and astrocytic defects have been described in the MECP2 mutant mouse (106). In vitro experiments indicate that the expression of astrocytic MECP2 is important for neuronal functions and disturbed in Mecp2 mutant mice (107–109). A broader involvement of astrocytes in the clinical phenotype of Rett Syndrome is suggested by rescue experiments in globally Mecp2-deficient mice which demonstrate that the re-expression of MECP2 preferentially in astrocytes improved not only respiratory abnormalities and significantly prolonged the lifespan, but it also improved locomotion, and anxiety levels (110). In an elegant study (111), it has been demonstrated that the calcium responses of ventral medullary astrocytes in Mecp2-KO mice were dramatically reduced when challenged with 10% CO2. The altered calcium response was reflected in a reduction of the sensitivity of brainstem astrocytes to changes in PCO₂/[H⁺]. These experiments are consistent with the notion that astrocytes play a critical role in the generation of the chemosensory response (112, 113), which involves the retrotrapezoid nucleus (RTN), a key region responsible for the CO₂ response (114). In addition to the evidence suggesting astrocytic involvement in Rett breathing disturbances, microglia have also shown to possibly contribute (115). However, as will be discussed in the next section, disturbances in neuromodulators and neurotransmitters also contribute to an altered chemosensory response.

The role of neurotransmitters and neuromodulators in Rett Syndrome

In Mecp2 KO mice, deficiencies were reported for gamma-Aminobutyric acid (GABA) (90, 116), noradrenaline (norepinephrine; (81)), serotonin (81, 117), and brain derived neurotrophic factor (37, 118). Disturbances in serotonin (5-HT) might have far-reaching consequences, given its important role in respiratory rhythmogenesis (119). Specifically, a deficit in 5-HT1A-mediated inhibition could contribute to the respiratory cycle irregularity seen in Rett Syndrome (120). In Rett Syndrome patients with confirmed MECP2 mutations, spinal fluid levels of a 5-HT metabolite are decreased (121), which may or may not be indicative of serotonergic alterations in the CNS. The strongest evidence for a disturbed CNS serotonin metabolism comes from animal models, which show low 5-HT levels in Mecp2 KO male mice (81, 122). The reduced levels of 5-HT seem to be due to the failure to synthesize serotonin (122). 5-HT plays an important role in the blunted CO₂ response in Rett Syndrome (95). Serotonergic neurons have been implicated in the hypercapnic response (123), and global deletion or acute inhibition of serotoninergic neurons in mice results in blunted hypercapnic responses (124, 125). Thus, in addition to the astrocytic deficiencies as described above (111), the serotonergic deficiencies likely contribute to the reduced CO₂ sensitivity in Mecp2 mice (82). Indeed, CO₂ sensitivity was corrected in the Mecp2 mutant mouse by selective blockade of serotonin reuptake (41). In addition to the role in the chemosensory response, serotonin is also critically involved in the generation of breathholds. A 5-HT1A/7 agonist has been shown to abolish apneas in wild-type mice (92) and in Mecp2 deficient female mice (34). Moreover, systemic administration of 5-HT1A agonists can rescue the apneas in Rett Syndrome (126). Levitt and co-workers elegantly demonstrated that a specific 5-HT1A agonist (F15599) selectively activates somatodendritic 5-HT1a autoreceptors and G protein coupled inwardly rectifying potassium (GIRK) channels on dorsal raphe neurons. This in turn leads to the activation of a slow inhibitory postsynaptic current. The same agonist led to a dose-dependent reduction in the apnea frequency. The duration of the individual apneas as well as the irregularity score of the respiratory frequency were also significantly reduced, both in male and female Mecp2 KO mice. It has been hypothesized that the activation of GIRK channels inhibits the activation of postinspiratory and late expiratory neurons mediating the breathhold, which likely involves the KF (126). The serotonergic effects observed in the mouse model led to a clinical trial (“STARS”) aimed at testing the efficiency of a 5HT1a receptor agonist, sarizotan, in treating respiratory dysfunction in Rett Syndrome (117, 127).

GABAergic mechanisms in the KF have also been implicated in Rett Syndrome (34). Substances that enhance GABAergic mechanisms can also reduce breathhold events and breathing irregularities and have been shown to restore chemosensitivity (34, 41). It has been hypothesized that several clinical symptoms of MECP2 are caused by developmental changes in the expression of NMDA and GABA receptor subunits (128, 129). By indirectly overexciting postinspiratory and late expiratory neurons, these synaptic mechanisms could lead to the breathholds (83, 90, 117, 130).

Breathholds, intermittent hypoxia and oxidative stress in Rett Syndrome

Seventy-five percent of the Rett Syndrome patient population develop breathholds before the age of 6 years, which is followed by hyperventilation 2-3 years later. As described above, disturbances in serotonergic and GABAergic mechanisms in the pons, in particular the KF region are likely responsible for the generation of the breathholding. The KF is part of the parabrachial nucleus, a region which has been implicated in anxiety and the sense of danger (131). The parabrachial nucleus also mediates the neuropathic response to pain (132) and fear (133). Thus, it is tempting to hypothesize that the parabrachial nucleus plays a critical role in linking the initiation of breathholds to the emotional aspects of the breathing disturbances in general, and the breathholds in particular (62). The activation of breathholds themselves cause intermittent hypoxia (81, 83, 134, 135), which could trigger a series of events. Exposure to intermittent hypoxia leads to overactive carotid bodies (136), which in turn might lead to the characteristic hyperventilation and a resulting hypocapnia (137). The hypocapnia may trigger a maladaptive response, given the altered serotonergic modulation and reduced ability of astrocytes to sense changes in PCO₂ /[H⁺] in Mecp2 mice (111). Intermittent hypoxia itself is known to desynchronize the central respiratory network which results in respiratory irregularities (138, 139) that are reminiscent to those seen in Mecp2 mice (81) and Rett patients (49). A vicious cycle can be triggered in which the desynchronization within the medullary respiratory network increases the probability for transmission failures to the hypoglossal nucleus, which in turn facilitates the appearance of obstructive sleep apneas (Fig. 3) (138–140) as has been described in Rett Syndrome (49). Intermittent hypoxia also leads to disturbances in sympathetic and parasympathetic control which contributes to dysautonomia (136, 141–145). The effects of intermittent hypoxia are largely mediated by reactive oxygen species (ROS) and oxidative stress (138, 139), which is further aggravated in Rett Syndrome, because of altered mitochondrial structure and enzyme activity: mitochondria are swollen, have altered respiratory rates and show proton leakage (146–150). Indeed, there is ample evidence for oxidative stress in Rett Syndrome (151–155). Targeting ROS with antioxidants is one therapeutic approach that could prevent some of the phenotypes associated with oxidative stress (29).

Breathing dysfunction and the coordination with swallowing.

Seventy percent of the Rett Syndrome patient population have difficulty coordinating breathing and swallowing, which results in dysphagia (156, 157). Several phenotypes of Rett Syndrome affect swallowing, including malocclusion (open bite) (156, 158), difficulty chewing (158–160), poor tongue mobility (156–158, 161), involuntary movements of the tongue (156, 158), poor bolus formation (160) and poor glossopharyngeal seal. Moreover, the pharyngeal cavity over distends (160), which requires more food/liquid to fill the pharyngeal cavity before the so called pharyngeal phase of swallowing can be initiated. There are also reports suggesting the absence of primary and secondary gastric motility and gastroesophageal reflux (157). Because of the complications associated with dysphagia, it is recommended to offer Rett Syndrome patients diets with thickened liquids and purees (159).

Both human Rett Syndrome and Mecp2 mice experience aerophagia (air swallowing) and malocclusion (158, 162, 163). Thus, swallow-breathing coordination in Rett Syndrome is weak and in some cases absent (161). Many Rett Syndrome patients are preferential mouth breathers (156) which probably contributes to the increased prevalence of aerophagia causing air bloat (162). According to the Natural history study, 47% of Rett parents also reported a history of air-swallowing during wakefulness (17). Although aerophagia is seen in other neurological and oropharyngeal diseases, it seems to be most common with Rett Syndrome (162), leading to painful distension of the stomach and even bowel perforation (162). Interestingly, the oral motor patterns of the tongue mirrored the motor patterns of the hands (160). In patients with hypotonia and normal tone, wringing of the hands was accompanied by abnormal motion of the tongue; and those patients with more rigidity/hypertonia whose hands lay motionless were also motionless in their tongue (160). These observations suggest that Rett Syndrome is characterized by complex imbalances that concurrently affect several rhythmogenic networks. Dysphagia is a significant, but understudied clinical presentation of Rett Syndrome, and pneumonia is one of the most frequently reported medical conditions (164) and the most common cause of death in Rett Syndrome according to one study (165). It is unknown whether pneumonia is the result of predisposed lung conditions or due to aspirations. However, according to some clinical reports it appears that aspiration has low occurrence in early stages of Rett Syndrome, but it is likely that the occurrence increases as the disorder progresses (160). In Mecp2-null mutant mice aspirations are frequent (64%) (163). Perhaps these aspirations were measured at more advanced states of disease progression. Alternatively, this difference is due to the fact that rodents are unable to cough, thus are more prone to aspirations (166). Fifty percent of Mecp2 mice exhibited lung inflammation and emphysema-like structural changes in the lung, which could be exaggerated by the aspirations (163).

Sleep and apneas in Rett Syndrome

In addition to the respiratory and integrated motor dysfunction observed during wakefulness, Rett patients display sleep disturbances and nocturnal respiratory abnormalities. The classic clinical presentation of EEG activity in Rett Syndrome is characterized by an abnormal slowing of background activity with epileptiform or sharp and spike waves, which declines as the disease progresses (167). Quantitative analysis of EEG activity has shown that Rett patients tend to spend less time in slow-wave-sleep throughout the night, however, display an increased delta power within each slow-wave sleep cycle (168). Additionally, sleep architecture among Rett Syndrome patients is characterized by decreased rapid eye movement (REM) sleep and increased non-REM (NREM) sleep, mainly the N3 stage (169) (See Figure 3). Alterations in sleep architecture among Rett syndrome is also accompanied by episodes of sleep disordered breathing (SDB). Several cohorts have reported the presence of SDB with a differing predominance of either obstructive or central sleep apnea among Rett patients (49, 169, 170). One of the most recent cohorts of 17 Rett Syndrome patients reported predominantly obstructive events in 76% of patients, of which 4 underwent previous adenotonsillectomy with incomplete resolution of their OSA, suggesting that adenotonsillar hypertrophy may not be the main mechanism of obstruction in these patients (169). This further suggests that a central mechanism may be largely contributing the occurrence of SDB in these patients.

Conclusion

Since the first description by Andreas Rett, much has been learned about Rett Syndrome. The identification of the gene that causes most of the Rett Syndrome cases in 2000 triggered a flurry of investigations to unravel the mechanisms leading to the various clinical presentations of Rett Syndrome. Among the many clinical manifestations, breathing disturbances are perhaps the most detrimental. Breathing disturbances are associated with autonomic dysfunctions that have been associated with sudden death. Dyscoordination with swallowing may contribute to pneumonia which is also a major clinical problem. Additionally, disturbances in sleep architecture and respiratory abnormalities during sleep contribute to the clinical manifestation observed in Rett.

The characterization of the neuronal mechanisms suggests that various cellular and neuromodulatory mechanisms are disturbed in Rett Syndrome. These mechanisms are complex and contribute directly and indirectly to the clinical presentations of Rett Syndrome. For example, ROS, triggered by the intermittent hypoxic events associated with the breathholds, may cause the dysautonomia. Promising animal research suggests that a number of pathologies can be reversed raising the hopes for clinical breakthroughs. Moreover, novel molecular and genetic techniques have allowed us to broaden our understanding of Rett Syndrome, which inspire novel clinical therapeutic strategies.