Sleep-disordered Breathing in Newborn Mice Heterozygous for the Transcription Factor Phox2b

    INSERM U676 Service de Reeanimation, Hpital Robert-Debree
    CNRS UMR 8542, Ecole Normale Supeerieure, Paris, France

    ABSTRACT

    Rationale: Central congenital hypoventilation syndrome (CCHS) is a rare autosomal dominant syndrome present from birth, and characterized by depressed ventilation during sleep. Heterozygous mutations of the homeobox gene Phox2b were recently found in a very high proportion of patients. Objectives: To determine whether newborn mice with heterozygous targeted deletion of the transcription factor Phox2b would display sleep-disordered breathing. Methods: We measured breathing pattern using whole-body plethysmography in wild-type and mutant 5-day-old mice, and we classified sleepeCwake states using nuchal EMG and behavioral scores. Results: We found that sleep apnea total time was approximately six times longer (8.9 ± 12 vs. 1.5 ± 2.2 seconds, p < 0.0015), and ventilation during active sleep was 21% lower (18.4 ± 5.1 vs. 23.3 ± 5.5 ml/g/second, p < 0.006) in mutant than in wild-type pups. During wakefulness, apnea time and ventilation were not significantly different between mutant and wild-type pups. Mutant and wild-type pups showed highly similar sleepeCwake states. Conclusion: Although their respiratory phenotype was much less severe than CCHS, the Phox2b+/eC mutant mice showed sleep-disordered breathing, which partially modeled the key feature of CCHS.

    Key Words: apnea; hypoventilation; Ondine syndrome

    Central congenital hypoventilation syndrome (CCHS, or Ondine's curse) is a rare autosomal dominant syndrome, present from birth and characterized by autonomic nervous system dysfunction, especially abnormal control of breathing during sleep, in the absence of primary neuromuscular or lung disease or an identifiable brainstem lesion. Breathing disorders show considerable interindividual variability, ranging in severity from relatively mild hypoventilation during quiet sleep, with adequate ventilation during wakefulness, to complete apnea during sleep and severe hypoventilation during wakefulness (1). Patients with CCHS have absent or markedly reduced ventilatory responses to hypercapnia and hypoxia, probably because of impairment of central integration of chemosensory afferents at the nucleus tractus solitarius level (2eC4) or central modulation of these responses at suprapontine and cerebellar levels (4eC6). All patients require mechanical support for sleep. The pathophysiologic mechanisms of CCHS are still largely unknown.

    The recent discovery of heterozygous mutations of the homeobox gene Phox2b leading to alanine expansion within the protein in a very high proportion of patients (2, 3, 7, 8) identified Phox2b as a key gene of CCHS. Phox2b is a master regulator of the noradrenergic phenotype and of all neuronal relays of the autonomic medullary reflex pathways (9), including peripheral chemoreceptors and their afferent visceral pathways (10).

    Mice heterozygous for the Phox2b mutation provide a promising approach to determine the pathophysiologic mechanisms of CCHS. Homozygous mutant mice lacking Phox2b die in utero around embryonic Day 14 (11). At least three neuronal types involved in chemosensitivity to hypercapnia and hypoxia are strictly dependent on Phox2b for their differentiation: the carotid body, the petrosal chemoreceptors that innervate it, and the nucleus tractus solitarius on which they project (10). We recently showed that newborn heterozygous mutant mice (i.e., with only one functional allele of Phox2b), which survive apparently normally, showed a blunted response to CO2 and augmented ventilatory decline in response to hypoxia (10). These results provided a partial mechanistic insight into CCHS. However, the mutant mice rapidly recovered within 10 days, and a key feature of CCHS, the fact that hypoventilation predominantly occurs during sleep, was not taken into account in our previous study. This aspect is highly relevant to assess Phox2b mutant mice as a valid model of CCHS.

    In this study, we hypothesized that genetic deficiency of Phox2b would lead to sleep-disordered breathing, in line with the main clinical feature of CCHS. To test this possibility, we classified for the first time in newborn mice the sleepeCwake states as wakefulness, active sleep, and quiet sleep based on nuchal EMG and behavioral indices. We determined how breathing varied as a function of sleepeCwake states in mutant and wild-type Phox2b 5-day-old mice. Postnatal Day 5 in mice corresponds approximately to 25 to 30 weeks' gestation in humans regarding the main aspects of brain development (12). This is a period of severe respiratory instability in preterm infants, who represent 20% of patients with CCHS (13).

    METHODS

    Animals

    The generation and genotyping of Phox2b mutant mice have been reported (14). Ten wild-type females crossed with mutant males gave birth to 22 heterozygous mutant mouse pups and 30 wild-type littermates, which were used as control pups. All the pups were tested at 5 days of age.

    Protocol

    Each pup was anesthetized with isoflurane according to a method recently developed to induce effective anesthesia (i.e., the abolition of the response to noxious stimuli) and rapid recovery (10 minutes) (15). Each pup was implanted with electrodes for EMG and placed inside the plethysmograph for 1 hour before recording to ascertain that the pup was familiarized with the environment. After adaptation, we recorded breathing, EMG, and behavioral scores for 15 minutes, and then weighed each pup and assessed its temperature by placing a copper-constantan thermocouple at skin level in the interscapular region (16).

    Whole-Body Flow Plethysmography

    Breath duration (TTOT, seconds), VT (e · geC1), ventilation E (calcu-lated as VT · TTOTeC1, and expressed in e · secondeC1 · geC1), and apneas (ventilatory pauses longer than twice the duration of the preceding breath) were measured noninvasively using whole-body flow barometric plethysmography as previously described (17). Apneas were determined using an automatic classification method recently developed (18). E was calculated on apnea-free periods.

    Nuchal EMG

    Two built-in Kynar hook electrodes (0.24 mm in diameter) were inserted through the skin into the nuchal muscle and secured with adhesive tape. We placed the pup in the supine position to easily observe motor twitches of the individual limbs and coordinated movements (19). An aluminum sheet, which served as a ground electrode, was secured to its abdomen. This sheet prevented righting while preserving limb movements. The EMG signals were collected at a sample rate of 500 Hz.

    Behavioral Scoring

    The experimenter visually recorded motor twitches and the coordinated movements by pressing one of two keys connected to the acquisition system (19). Motor twitches were defined as phasic, rapid, and independent movements of one limb or the tail. Coordinated movements were defined as sustained motor activity of the limbs or the head.

    SleepeCWake States

    Wakefulness, active sleep, and quiet sleep were determined off-line. An experienced analyzer examined the entire EMG recording to visually estimate the magnitude of changes of EMG and discriminate between silent, low, and high EMG. Then, the experimenter classified the entire signal as active sleep, quiet sleep, wakefulness, or undetermined state. Active sleep was characterized by silent nuchal EMG without coordinated movements, and possible occurrence of motor twitches (19eC21). Quiet sleep was characterized by low nuchal EMG not accompanied by motor twitches or movements. Wakefulness was characterized by high EMG activity associated with coordinated movements. Bouts of recordings that did not fit with any of these three definitions were classified as undetermined states.

    Statistics

    Breathing variables were averaged over each sleepeCwake state (wakefulness, active sleep, quiet sleep), and then subjected to analyses of variance (22), with genotype as the between-subject factor and sleepeCwake states as the within-subject factor.

    RESULTS

    Weight and Temperature

    Mutant pups weighed significantly less than wild-type pups (2.9 ± 0.5 vs. 3.3 ± 0.6 g, respectively; p < 0.002), and had lower temperatures (32.6 ± 0.6 vs. 33.0 ± 0.5°C, respectively; p < 0.008). Temperature was not significantly related to body weight (neither as a main effect nor in interaction with genotype).

    Sleep Duration and Sleep States

    Recovery from anesthesia was similar in mutant and wild-type pups (additional details on the method and the results of this experiment are provided in an online supplement).

    After classification into wakefulness, active sleep, or quiet sleep, the total durations of undetermined states were very small in both genotype groups (mean values in mutant pups, 1.6 ± 5.1 seconds; wild-type, 1.0 ± 2.3 seconds; difference, not significant [NS]). Illustrative examples of EMG, behavioral, and respiratory tracings are shown in Figure 1.

    SleepeCwake states were very similar in mutant and wild-type pups (Figure 2). The proportion of sleep time was 78 ± 14% of total recording time in mutant pups versus 75 ± 13% in wild-type pups (difference, NS). The proportion of active sleep was 80 ± 12% of total sleep time in mutants versus 79 ± 18% in wild-type pups (difference, NS). Thus, active sleep was predominant when compared with wakefulness and quiet sleep in both groups (Figure 2). There was no difference in number, mean dura- tion of bouts, and total duration for each state between mutant and wild-type pups (all main effects for genotype and state-by-genotype interactions, NS; Figure 2).

    Apneas and Ventilation

    Large proportions of respiratory signal free from artifacts were obtained in both mutant and wild-type pups (87 ± 7 vs. 92 ± 5%). This proportion was slightly lower in mutant pups (p < 0.01), probably because of larger respiratory instability.

    Phox2b+/eC pups had a longer apnea time (as percentage of the total duration of sleepeCwake states) than Phox2b+/+ pups (main effect for genotype, p < 0.007; Figure 3). This difference was entirely because of their longer apnea time during sleep (genotype-by-state interaction, p < 0.001; and partial comparisons between genotype groups during active sleep, p < 0.003, during quiet sleep, p < 0.027, and during wakefulness, NS; Figure 3). In Phox2B+/eC pups, the apnea times (in seconds) during wakefulness, active sleep, and quiet sleep were as follows: 0.00 ± 0.00, 7.63 ± 10.70, and 1.26 ± 2.80, respectively. In Phox2B+/+ pups, the apnea times (in seconds) were 0.05 ± 0.20, 1.29 ± 1.80, and 0.27 ± 0.67. The analysis of absolute apnea time yielded similar statistical results as those of percentage of time (main effect for genotype, p < 0.002; genotype-by-state interaction, p < 0.0003; and partial comparisons between genotype groups, during active sleep, p < 0.0023, during quiet sleep, p < 0.066, and during wakefulness, NS).

    Interindividual variability of apnea time was much larger in mutant than in wild-type pups (Figures 3 and 4). Furthermore, six mutant pups (of 22) showed apnea times outside the range of wild-type values. To examine whether this variability reflected a nonspecific developmental delay, we searched for relationships between body weight and apnea time. Analysis of covariance with weight as a cofactor and genotype as a between-subject factor revealed that weight was weakly correlated with apnea time in mutant (r = 0.37) but not in wild-type pups (genotype-by-weight interaction, p < 0.046; Figure 4). Apnea time was not significantly related to body temperature.

    Compared with wakefulness values, the mean E decline during sleep (both states included) was 44 ± 17% in mutant pups and 30 ± 18% in wild-type pups (p < 0.008). The lower E of Phox2b+/eC pups was because of a smaller E during active sleep (genotype-by-state interaction, p < 0.026; partial comparison mutant vs. wild-type pups for active sleep, p < 0.002; other genotype-related comparisons, NS). The lower E of mutant pups compared with wild-type pups was caused by a longer TTOT (0.30 vs. 0.26 seconds, respectively; p < 0.018); the difference for VT was not significant. As a rule, the pups that showed large E decreases from wakefulness to active sleep were those with longer apnea times in both groups (Spearman's rank correlation test, p < 0.015 and p < 0.005, respectively).

    To examine whether the E decline during sleep was related to developmental delay or to impaired metabolism, we conducted separate analyses of covariance with weight and body temperature as cofactors and genotype as a between-subject factor. We found that weight had no significant effects (Figure 5). The E decline was significantly related to body temperature (p = 0.0001), although this effect was not significantly different in mutant and wild-type pups (genotype-by-temperature interaction, NS; Figure 5).

    Then, we analyzed breathing variables in weight-matched groups of mutant and wild-type pups, using the "nearest available" method (23), with a maximal difference of 0.1 g (Table 1). Matching on weight produced two subsamples of 16 mutant and wild-type pups, which had closely similar body temperatures (Table 1). SleepeCwake states were practically identical in the two subgroups. Phox2b+/eC pups had a longer apnea time than weight-matched Phox2b+/+ pups (main effect for genotype, p < 0.035) because of their longer apnea time during active sleep (genotype-by-state interaction, p < 0.023; and partial comparisons during active sleep, p < 0.029; Table 1). Finally, we matched mutant and wild-type pups on body temperature (maximal difference, 0.2°C). Phox2b+/eC pups had a longer apnea time than temperature-matched Phox2b+/+ pups (main effect for genotype, p < 0.018) because of their longer apnea time during active sleep (genotype-by-state interaction, p < 0.011; and partial comparisons during active sleep, p < 0.024; Table 1). Taken together, these results showed that sleep-related disorders in mutant pups were not secondary to the lower weights and body temperature.

    The coefficients of variation of E were significantly higher in mutant than in wild-type pups in all sleepeCwake states (wakefulness, 41 ± 6 vs. 35 ± 9%, p < 0.009; active sleep, 43 ± 13 vs. 34 ± 15%, p < 0.031; and quiet sleep, 47 ± 14 vs. 35 ± 15%, p < 0.005, respectively). This effect was caused by the greater variability of TTOT in mutant than in wild-type pups (wakefulness, 39 ± 6 vs. 35 ± 9%, p < 0.032; active sleep, 23 ± 5 vs. 19 ± 4%, p < 0.001; quiet sleep, 26 ± 7 vs. 20 ± 6%, p < 0.002). Wakefulness was associated with increased variability of TTOT (but not E) compared with active and quiet sleep levels (p < 0.001 for both comparisons) irrespective of genotype (the differences in E or TTOT variability between active and quiet sleep were not significant). Apneas were practically absent in both groups during wakefulness (Figure 3).

    Finally, we further examined the possibility that the differences in breathing variables between mutant and wild-type pups were caused by delayed recovery from anesthesia in mutant pups (not revealed by behavioral tests; see Figure E1 in the online supplement). To check this possibility, we analyzed exploratory data collected in a few pups (five mutant and eight wild-type) over 15 minutes after the study period. The difference in E decline between mutant and wild-type mice was similar to previous results in the overall sample (43 ± 24 vs. 21 ± 25%). The total apnea time (all during active sleep) was 5.47 ± 6.75 versus 0.66 ± 1.50 seconds. Both results constitute further evidence that sleep-related respiratory disorders were not caused by residual effects of anesthesia.

    DISCUSSION

    The aim of this study was to determine whether newborn mice with targeted deletion of the transcription factor Phox2b, which was found to be mutated in patients with CCHS, would display sleep-disordered breathing. We found that mutant Phox2b+/eC newborn mice showed longer apnea time during sleep than their Phox2b+/+ wild-type littermates, and lower ventilation during active sleep. There was no difference in sleepeCwake state durations between mutant and wild-type pups. Thus, Phox2b+/eC mice showed sleep-related respiratory disorders, which partially model CCHS.

    SleepeCWake States

    The classification of states into wakefulness, active sleep, and quiet sleep has not been previously described in newborn mice. The method for sleepeCwake states classification used in our investigation was derived from previous studies in newborn mammals (20), including rats (19, 21, 24). In newborn rats, nuchal muscle tone, coordinated movements, and motor twitches have proved reliable for identifying sleepeCwake states between postnatal day 2 and postnatal day 8 (21). Using these criteria, we found that the proportions of sleep time with respect to recording time and the proportion of active sleep with respect to sleep time were closely similar to those reported in 2- and 8-day-old rats (21). In particular, active sleep was predominant, reaching approximately 80% of sleep time, which is close to the values reported in human preterm infants (25). These similarities support the validity of our sleepeCwake classification method.

    One limitation of our study is that we placed the pups in the supine position to observe motor twitches and coordinated movements while minimizing movement artifacts on the EMG (19). A recent study in newborn rats showed that supine position did not affect sleepeCwake states (21). On the other hand, restraining the animals, even loosely, may have affected breathing pattern, as shown in adult mice (26). Thus, the differences in breathing pattern between mutant and wild-type pups in the supine position in our study may be partly ascribable to different responses to restraint.

    In this study, the classification of sleepeCwake states was done before knowledge of genotypes, thus precluding any bias in genotype-group comparisons. The fact that sleepeCwake states were normal in mutant pups suggested that apparently the damage was not shared between breathing control and sleepeCwake states mechanisms. However, we cannot rule out that qualitative differences within a given sleepeCwake state existed between mutant and wild-type pups.

    Sleep-disordered Breathing in Phox2b+/eC Pups

    The absolute values of VT and E obtained by whole-body plethysmography may be biased by possible genotype-related differences in body temperature inside the plethysmograph (not measured). However, the present results are mainly based on TTOT and apnea values, which are validly measured by whole-body plethysmography. TTOT variability was greater in mutant than in wild-type pups irrespective of sleep-wake states, whereas apnea time was longer in mutant pups during sleep only. Thus, breathing variability and apnea durations provided complementary information. The temperature of the plethysmograph chamber was maintained at 32.8°C. The temperatures measured by inserting a probe in several litters of newborn mice in contact with the mother were 32°C (32). Thus, it is unlikely that the pups suffered from thermal stress.

    Sleep apnea time was longer in mutant pups, and ventilation during active sleep was approximately 20% lower in mutant than in wild-type pups. These impairments may be caused by the decrease of the tonic drive to breathe provided by chemosensitive sites acting predominantly during sleep (27, 28). Central chemoreceptors are distributed at many locations within the brainstem (29). In rats, the CO2-sensitive neurons of the caudal medullar raphe act during sleep only (28) and those of the rostral nucleus tractus solitarius are more effective during sleep than wakefulness (27). Other CO2-sensitive sites (locus ceruleus and area postrema [30, 31]) and the afferent pathways from carotid bodies, which contribute to CO2 sensitivity, also depend on Phox2b for their development (10). However, the activity of these sites has not been studied with respect to sleepeCwake specialization.

    In addition to respiratory disorders, we found that Phox2b+/eC mutant pups weighed less and had lower temperatures than their wild-type littermates. In mutant pups, sleep-related respiratory disorders and body temperature were correlated. These abnormalities, and the dilated pupils of adult Phox2b+/eC mice described by others (33), suggest a widespread autonomic disorder, with variable severity among individuals, as reported in patients with CCHS (5, 6, 13). This spectrum of autonomic symptoms is consistent with the broad range of autonomic nervous system targets of mutations affecting Phox2b, which is the master regulator of the noradrenergic phenotype and of all neuronal relays of the autonomic medullary reflex pathways (9).

    Both apnea time and the E decline during sleep were weakly correlated with body weight and temperature in mutant pups. However, mutant pups had longer sleep apnea times than weight- and temperature-matched wild-type pups (a similar trend was observed for the E decline), establishing that sleep-related disorders in mutant pups were not secondary to the lower weights and body temperature.

    We did not investigate whether rescue of the respiratory disorders in mutant pups was feasible (10). In previous studies, Phox2beC/eC embryos died shortly after midgestation but could be rescued by maternal treatment with noradrenalin agonists throughout gestation (34). The possibility (not studied here) that the abnormal respiratory phenotype of mutant Phox2b+/eC pups may be amenable to rescue strategies is of considerable clinical interest.

    Phox2b+/eC Newborn Mice: A Model for CCHS

    The enhanced sleep apneas and the smaller E during sleep, the smaller weights (45% patients with CCHS have developmental delays [13]), and the lower temperature (one of the associated symptoms of CCHS [35]), in addition to diminished ventilatory response to hypercapnia (10) are akin to CCHS. Furthermore, the large interindividual variability in respiratory disorders reported here is reminiscent of the large interindividual variability of respiratory disorders in patients with CCHS (1).

    On the other hand, the respiratory phenotype of Phox2b+/eC pups departed from CCHS in several aspects. First, this phenotype was much less severe than CCHS. This finding is not surprising, considering that Phox2b mutant mice survive and are fertile, whereas the outcome of patients with CCHS is extremely severe unless they receive lifelong ventilatory support during sleep. Second, the ventilatory impairments in mutant pups were more pronounced during active than quiet sleep, whereas the opposite pattern is found in patients with CCHS. Possibly, the small proportion of quiet sleep, which was because of immaturity of sleep organization in 5-day-old pups, reduced the accuracy of ventilation estimates during this stage. The brevity of the present recordings contrasts with the prolonged polysomnographic recordings obtained in patients with CCHS (36). More than 20% of patients with CCHS were born prematurely (13). However, polysomnographic recordings in these patients were performed at later stages of sleep development, compared with the present study. How sleepeCwake states affect breathing in preterm infants with CCHS early after birth has not been described and cannot be inferred from previous studies. For example, a full-term patient with CCHS monitored at 6 weeks of age displayed respiratory impairments only during periods of delta-wave sleep, which does not exist at birth (37). The lack of polysomnographic recordings early after birth in preterm patients with CCHS is an obstacle to comparisons with present data.

    Other differences have been noted between the phenotype of Phox2b heterozygote mice and CCHS. Cross and coworkers (33) recently reported that mutant mice have dilated pupils, whereas ocular abnormalities in patients with CCHS are, for the most part, constricted pupils (13). At this stage, we cannot exclude that these discrepancies stem from functional differences between the Phox2b targeted mutation in mice (which is clearly a null mutation) and the alanine expansion or homeobox distal mutations found in the Phox2b gene of patients with CCHS (13, 33, 38, 39).

    Conclusions

    Genetic deficiency of Phox2b induced sleep apneas and hypoventilation in newborn mice, thus providing the first transgenic mouse model of sleep-disordered breathing of infancy. The respiratory phenotype of Phox2b+/eC newborn mice, together with their reduced weights and temperature, and their normal organization of sleep and wakefulness are akin to CCHS. Thus, Phox2b+/eC mice showed sleep-related respiratory disorders that partially model CCHS.

    Acknowledgments

    The authors thank Guy Vardon, for developing the plethysmographic system; Boris Matrot, for his invaluable contribution to signal-processing algorithms; and Jolle Adrien (INSERM U288) for her helpful advice regarding the EMG studies.

    This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org

    REFERENCES

    Gozal D. Congenital central hypoventilation syndrome: an update. Pediatr Pulmonol 1998;26:273eC282.

    Weese-Mayer DE, Berry-Kravis EM, Zhou L, Maher BS, Silvestri JM, Curran ME, Marazita ML. Idiopathic congenital central hypoventilation syndrome: analysis of genes pertinent to early autonomic nervous system embryologic development and identification of mutations in PHOX2b. Am J Med Genet 2003;123A:267eC278.

    Weese-Mayer D, Berry-Kravis E. Genetics of congenital central hypoventilation syndrome. Am J Respir Critl Care Med 2004;170:16eC21.

    Macey PM, Valderama C, Kim AH, Woo MA, Gozal D, Keens TG, Harper RK, Harper RM. Temporal trends of cardiac and respiratory responses to ventilatory challenges in congenital central hypoventilation syndrome. Pediatr Res 2004;55:953eC959.

    Macey PM, Woo MA, Macey KE, Keens TG, Saeed MM, Alger JR, Harper RK. Hypoxia reveals posterior thalamic, cerebellar, midbrain and limbic deficits in congenital central hypoventilation syndrome. J Appl Physiol 2005;98:958eC969.

    Harper RM, Macey PM, Woo MA, Macey KE, Keens TG, Gozal D, Alger JR. Hypercapnic exposure in congenital central hypoventilation syndrome reveals central nervous system respiratory control mechanisms. J Neurosci 2005;93:1647eC1658.

    Amiel J, Laudier B, Attie-Bitach T, Trang H, de Pontual L, Gener B, Trochet D, Etchevers H, Ray P, Simonneau M, et al. Polyalanine expansion and frameshift mutations of the paired-like homeobox gene PHOX2B in congenital central hypoventilation syndrome. Nat Genet 2003;33:459eC461.

    Matera I, Bachetti T, Puppo F, Di Duca M, Morandi F, Casiraghi G, Cilio M, Hennekam R, Hofstra R, Schber J, et al. PHOX2B mutations and polyalanine expansions correlate with the severity of the respiratory phenotype and associated symptoms in both congenital and late onset central hypoventilation syndrome. J Med Genet 2004;41:373eC380.

    Pattyn A, Morin X, Cremer H, Goridis C, Brunet JF. Expression and interactions of the two closely related homeobox genes Phox2a and Phox2b during neurogenesis. Development 1997;124:4065eC4075.

    Dauger S, Pattyn A, Lofaso F, Gaultier C, Goridis C, Gallego J, Brunet JF. Phox2b controls the development of peripheral chemoreceptors and afferent visceral pathways. Development 2003;130:6635eC6642.

    Pattyn A, Hirsch M, Goridis C, Brunet JF. Control of hindbrain motor neuron differentiation by the homeobox gene Phox2b. Development 2000;127:1349eC1358.

    Marret S, Mukendi R, Gadisseux JF, Gressens P, Evrard P. Effect of ibotenate on brain development: an excitotoxic mouse model of microgyria and posthypoxic-like lesions. J Neuropathol Exp Neurol 1995;54:358eC370.

    Vanderlaan M, Holbrook CR, Wang M, Tuell A, Gozal D. Epidemiologic survey of 196 patients with congenital central hypoventilation syndrome. Pediatr Pulmonol 2004;37:217eC229.

    Pattyn A, Morin X, Cremer H, Goridis C, Brunet JF. The homeobox gene Phox2b is essential for the development of autonomic neural crest derivatives. Nature 1999;399:366eC370.

    Drobac E, Durand E, Laudenbach V, Mantz J, Gallego J. A simple method for short-term controlled anesthesia in newborn mice. Physiol Behav 2004;82:279eC283.

    Blumberg MS, Sokoloff G. Thermoregulatory competence and behavioral expression in the young of altricial species: revisited. Dev Psychobiol 1998;33:107eC123.

    Dauger S, Durand E, Cohen G, Lagercrantz H, Changeux JP, Gaultier C, Gallego J. Control of breathing in newborn mice lacking the beta-2 nAChR subunit. Acta Physiol Scand 2004;182:205eC212.

    Matrot M, Durand E, Dauger S, Vardon G, Gaultier C, Gallego J. Automatic classification of activity and apneas using whole-body plethysmography in newborn mice. J Appl Physiol 2004;98:365eC370.

    Karlsson KA, Blumberg MS. The union of the state: myoclonic twitching is coupled with nuchal muscle atonia in infant rats. Behav Neurosci 2002;116:912eC917.

    Jouvet-Mounier D, Astic L, Lacote D. Ontogenesis of the states of sleep in rat, cat, and guinea pig during the first postnatal month. Dev Psychobiol 1970;2:216eC239.

    Karlsson KA, Kreider JC, Blumberg MS. Hypothalamic contribution to sleep-wake cycle development. Neuroscience 2004;123:575eC582.

    Keselman HJ, Keselman JC. The analysis of repeated measures designs in medical research. Stat Med 1984;3:185eC195.

    Cologne JB, Shibata Y. Optimal case-control matching in practice. Epidemiology 1995;6:271eC275.

    Gramsbergen A, Schwartze P, Prechtl HF. The postnatal development of behavioral states in the rat. Dev Psychobiol 1970;3:267eC280.

    Lehtonen L, Martin RJ. Ontogeny of sleep and awake states in relation to breathing in preterm infants. Semin Neonatol 2004;9:229eC238.

    Dauger S, Nsegbe E, Vardon G, Gaultier C, Gallego J. The effects of restraint on ventilatory responses to hypercapnia and hypoxia in adult mice. Respir Physiol 1998;112:215eC225.

    Nattie EE, Li A. CO2 dialysis in nucleus tractus solitarius region of rat increases ventilation in sleep and wakefulness. J Appl Physiol 2002;92:2119eC2130.

    Nattie EE, Li A. CO2 dialysis in the medullary raphe of the rat increases ventilation in sleep. J Appl Physiol 2001;90:1247eC1257.

    Feldman JL, Mitchell GS, Nattie EE. Breathing: rhythmicity, plasticity, chemosensitivity. Annu Rev Neurosci 2003;26:239eC266.

    Nattie EE. Central chemosensitivity, sleep, and wakefulness. Respir Physiol 2001;129:257eC268.

    Ruggiero DA, Gootman PM, Ingenito S, Wong C, Gootman N, Sica AL. The area postrema of newborn swine is activated by hypercapnia: relevance to sudden infant death syndrome J Auton Nerv Syst 1999;76:167eC175.

    Renolleau S, Dauger S, Autret F, Vardon G, Gaultier C, Gallego J. Maturation of baseline breathing and of hypercapnic and hypoxic ventilatory responses in newborn mice. Am J Physiol Regul Integr Comp Physiol 2001;281:R1746eCR1753.

    Cross SH, Morgan JE, Pattyn A, West K, McKie L, Hart A, Thaung C, Brunet JF, Jackson IJ. Haploinsufficiency for Phox2b in mice causes dilated pupils and atrophy of the ciliary ganglion: mechanistic insights into human congenital central hypoventilation syndrome. Hum Mol Genet 2004;13:1433eC1439.

    Pattyn A, Goridis C, Brunet JF. Specification of the central noradrenergic phenotype by the homeobox gene Phox2b. Mol Cell Neurosci 2000;15:235eC243.

    American Thoracic Society. Idiopathic congenital central hypoventilation syndrome. diagnosis and management. Am J Respir Crit Care Med 1999;160:368eC373.

    Takahashi R, Kakizawa H, Itou T, Yamada M, Nakae S. Peculiar respiratory response observed during sleep-onset REM sleep of an infant with Ondine's curse. Neuropediatrics 2000;31:269eC272.

    Guilleminault C, McQuitty J, Ariagno RL, Challamel MJ, Korobkin R, McClead RE Jr. Congenital central alveolar hypoventilation syndrome in six infants. Pediatrics 1982;70:684eC694.

    de Pontual L, Nepote V, Attie-Bitach T, Al Halabiah H, Trang H, Elghouzzi V, Levacher B, Benihoud K, Auge J, Faure C, et al. Noradrenergic neuronal development is impaired by mutation of the proneural HASH-1 gene in congenital central hypoventilation syndrome (Ondine's curse). Hum Mol Genet 2003;12:3173eC3180.

    Sasaki A, Kanai M, Kijima K, Akaba K, Hashimoto M, Hasegawa H, Otaki S, Koizumi T, Kusuda S, Ogawa Y, et al. Molecular analysis of congenital central hypoventilation syndrome. Hum Genet 2003;114:22eC26.