Published ahead of print on December 13, 2007, doi:10.1164/rccm.200707-1079OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200707-1079OC
PHOX2B Germline and Somatic Mutations in Late-Onset Central Hypoventilation Syndrome1 INSERM U781 and Département de Génétique, Hôpital Necker-Enfants Malades, Université Paris 5–René Descartes, Paris, France; 2 Université Pierre et Marie Curie–Paris 6, UPRES EA2397, Service Central d'Explorations Fonctionnelles Respiratoires, Groupe Hospitalier Pitié-Salpêtrière, Paris, France; 3 Division of Pediatric Sleep Medicine, Department of Pediatrics, and Kosair Children's Hospital Research Institute, University of Louisville, Louisville, Kentucky; 4 Service de Physiologie, INSERM E9935, Hôpital Robert Debré, Université Paris VII, Paris, France; and 5 Service de Neuropédiatrie, Hôpital Bicêtre, Bicêtre, France Correspondence and requests for reprints should be addressed to Jeanne Amiel, M.D., Ph.D., Département de Génétique, Hôpital Necker–Enfants Malades 149, rue de Sèvres, 75743 Paris Cedex 15, France. E-mail: amiel{at}necker.fr
Rationale: Late-onset central hypoventilation syndrome (LO-CHS) is a rare disorder that may manifest as early as infancy or as late as during adulthood. The potential overlap of LO-CHS with congenital CHS is under debate, even though both disorders can result from heterozygous PHOX2B gene mutations. Objectives: To characterize the PHOX2B status in a series of 25 patients with LO-CHS referred from 3 months of age to adulthood. Whenever a PHOX2B mutation was identified, we ascertained its germline or somatic origin in both patients with LO-CHS and in 15 parents of probands with congenital CHS found to harbor a PHOX2B mutation. Methods: The PHOX2B gene was analyzed by direct DNA sequencing and origin of the mutation evaluated by fluorescent PCR. Measurements and Main Results: We have identified a heterozygous PHOX2B gene mutation in 17 of 25 patients with LO-CHS. The far most frequent mutation results in a germline +5 alanine expansion in the series of 20 alanines (15 cases) that show incomplete penetrance and variable expressivity, possibly resulting from combined environmental and genetic factors. PHOX2B frameshift and missense mutations have also been identified in patients with LO-CHS. Importantly, one parent found to harbor a somatic mosaic for a +8 alanine expansion developed alveolar hypoventilation in his 40s. Conclusions: These data indicate that PHOX2B gene mutations should be systematically examined in any adult with unexplained central hypoventilation and raise the question of follow-up for apparently healthy parents found to harbor a somatic mosaic for the PHOX2B mutation identified in their child.
Key Words: late-onset central hypoventilation • PHOX2B • somatic mosaic
Central hypoventilation syndrome (CHS) is defined as the presence of persistent alveolar hypoventilation and/or apnea during sleep and impaired ventilatory responses to hypercapnia. Initial cases of central hypoventilation were those secondary to traumatic, vascular, infectious, or tumor events (1). In 1970, Mellins and collaborators reported the first case of idiopathic congenital CHS (C-CHS; Ondine's curse; MIM209880) (2). Since this initial report, it has become apparent that C-CHS is associated with broader impairments of the autonomic nervous system, rather than being restricted to impaired respiratory control, and that predisposition to Hirschsprung disease (HSCR) occurred in 20% of the cases and predisposition to tumors of the sympathetic nervous system (TSNS) in 5 to 10% of the cases (3, 4). The spectrum of autonomic manifestations is wide and encompasses ocular, gastroesophageal, and cardiac involvement (5, 6). As experience with the diagnosis and treatment of C-CHS accumulated, patients with idiopathic late-onset CHS (LO-CHS) were reported, in whom the onset of symptoms would occur beyond the neonatal period and as late as during adulthood. There is, however, considerable debate as to whether LO-CHS represents a separate disease entity from C-CHS, or whether it should be included as part of the spectrum of C-CHS. As a consequence of this debate, a subset of patients presenting with hypothalamic dysfunction (HD) before onset of central hypoventilation have been identified and termed LO-CHS–HD (7–9).
The paired-like homeobox gene (PHOX2B) encodes a key transcription factor in the development of the central and peripheral autonomic nervous system (10). Through a candidate gene approach, we previously identified PHOX2B as the gene underlying C-CHS, with a de novo heterozygous mutation being present in the vast majority of patients (11). The most prevalent mutation is an expansion ranging from +5 to +13 alanines in a stretch of 20 alanines ( In the present study, we have focused our genetic analysis on a series of 25 patients presenting with isolated LO-CHS and on 15 additional subjects who were the parents of patients C-CHS, and who carried a PHOX2B mutation. Our findings show that an expansion of five alanines is by far the most frequent in LO-CHS, but other mutations can also be ascribed to later onset of ventilatory symptoms. In addition, individuals who carry a PHOX2B mutation in a mosaic state may become symptomatic over time. Some of the results of these studies have been previously reported in the form of an abstract (19).
Patients Subjects enrolled in the study were divided into two groups: 25 cases with a diagnosis of isolated LO-CHS (among whom 7 patients were previously reported [14, 20]) and 15 carrier parents of a PHOX2B mutation identified in the index case with C-CHS or LO-CHS (among whom 10 were previously quoted [14]). Patients diagnosed with LO-CHS–HD have been excluded from the present study. Subjects with LO-CHS (13 females and 12 males) were recruited in France (n = 13), the United States (n = 5), Australia (n = 2), Poland (n = 2), England (n = 1), Spain (n = 1), and Israel (n = 1). The clinical data are summarized in Table 1. Inclusion criteria for patients with LO-CHS were as follows: (1) spontaneous and efficient ventilation in the newborn period (i.e., first 28 d of life); (2) absence of any identifiable neuromuscular, cardiac, or pulmonary cause; and (3) no hypothalamic dysfunction before onset of ventilatory symptoms. Therefore, patients presenting hyperphagia, obesity, and/or endocrinopathy were excluded. Age at onset varied from 3 months to 55 years. A large majority had their LO-CHS diagnosed during childhood (n = 18), with eight receiving their diagnosis before the age of 1 year; three patients had their LO-CHS diagnosed during adolescence and the remaining four patients received their diagnosis in adulthood. Among the putative triggering events that prompted their referral and diagnosis, upper airway infection (2 cases), lower airway respiratory infection (6 cases), and anesthesia (2 cases) were identified. Pulmonary hypertension and right heart hypertrophy indicating chronic hypoventilation before the acute event were observed in eight cases. Several patients presented episodes of respiratory distress at birth or during the first few weeks of life for which no cause was identified and with spontaneous resolution. Some patients (n = 9) experienced severe pulmonary infections requiring mechanical ventilation in the intensive care unit, and were able to discontinue ventilatory support without problems. In particular, patient O237, who received a diagnosis at age 17 years, presented respiratory failure at age 14 months and 7 years and received oxygen therapy at that time. Patient O274 received a diagnosis at 12 years, and was known for his unusual capabilities in breath-holding diving in the swimming pool. In all cases, after the diagnosis of LO-CHS was established, ventilatory support was necessary during sleep only and gas exchange remained stable and within normal limits during wakefulness. Ocular anomalies, such as strabismus and pupillary contraction anomalies, were reported in four cases. HSCR and TSNS were not reported. Of note, patient O266 had a brother who died of sudden infant death syndrome (SIDS) in the first year of life.
The second group of subjects comprised 15 parents of C-CHS cases carrying the PHOX2B mutation identified in their child (8 males, 7 females). Their ages ranged from 25 to 58 years (Table 2). Most were apparently healthy (14/15), although one developed sleep apnea at age 40 years (subject S194f), for which he required nighttime ventilatory intervention with continuous positive airway pressure.
Blood samples were obtained with informed consent and DNA was extracted according to standard protocols.
DNA Sequencing
Analysis of Putative Somatic Mosaicism
PHOX2B Mutation Screening in Idiopathic LO-CHS Cases We performed PHOX2B direct sequencing in the 25 idiopathic LO-CHS cases and found 17 heterozygous PHOX2B mutations. Fifteen patients harbored a mutation leading to a heterozygous expansion of +5 alanines. Among those patients, one 55-year-old woman (patient O44GM) was referred for testing after developing respiratory arrest after anesthesia. She had a son and a granddaughter with C-CHS and harboring a germline +5 alanine expansion (14). Similarly, we identified a +5 alanine expansion in a female with LO-CHS at 3 months of age inherited from her father who presented LO-CHS at the age of 25 years (patients O299 and O299f). Other types of heterozygous mutations were found. A heterozygous nucleotide change occurring in exon 2 and leading to the change of a highly conserved alanine to glutamic acid at position 140 within the homeodomain (c.419C>A; A140E) was identified in a female subject who required tracheotomy placement at 8 months of age, and who presented evidence of sustained hypercapnia before tracheotomy (O201; Table 1). Finally, we identified a heterozygous G deletion in exon 3 at position c.692 of the coding sequence, leading to a 307 amino acid truncated protein (c.692delG, p.Gly231AlafsX78), in a 1.5-year-old boy presenting with sleep apnea and hypercapnia. In seven of the nine cases in whom parents (only the mother in case O188) were available for study, the mutation occurred de novo. The +5 alanine expansion was inherited from an asymptomatic father in one case, and from a father with LO-CHS in another case (Table 1; patients O106 and O299).
Somatic Mosaicism versus Incomplete Penetrance
Conversely, all LO-CHS cases carrying a + 5 alanine expansion reported here seem to harbor a germinal mutation, because the mutant versus wild-type ratio equals 1 in their leukocytes. This is also the case for two asymptomatic parents harboring a +5 alanine expansion who remain asymptomatic at ages 43 and 39 years, respectively (patients O102m and O106f; Figure 1D and data not shown).
The distinction between C-CHS and LO-CHS is difficult and not well defined, and no consensual definition of LO-CHS is available to date. We propose that patients who are diagnosed with CHS and who presented with the ability to sustain autonomous breathing in the neonatal period be classified as LO-CHS. We report a PHOX2B mutation in 17 of 25 patients with such clinical presentation. However, we should also point out that some patients with LO-CHS presented ventilatory abnormalities in the neonatal period, which resolved spontaneously (e.g., cases O234 and O235 in our series; Table 1). Although we cannot exclude a disease-causing mutation outside of the coding sequence and exon–intron boundaries at the PHOX2B locus, genetic heterogeneity is likely in the remaining 8 of 25 patients. Table 1 lists additional features identified in these patients. Some may be useful for future delineation of distinct entities in the group of patients with LO-CHS not ascribed to PHOX2B gene mutation. In this study, we conclusively demonstrate that, similar to C-CHS, various PHOX2B mutations can lead to LO-CHS (i.e., polyalanine expansions, missense mutations, and frameshift mutations). This is in contrast with all LO-CHS cases with a PHOX2B mutation reported thus far in the literature in whom only a + 5 alanine expansion was identified (13, 16–18, 20). Although we do not know whether the published LO-CHS cases harbor a germinal mutation or a somatic mosaicism, the present study shows that all cases with LO-CHS reported herein who had a +5 alanine expansion harbored a germinal mutation, and that this was also the case in two asymptomatic parents. Interestingly, one of these parents (patient 0106f) has a child with LO-CHS. These observations support the hypothesis that a +5 alanine expansion can remain incompletely penetrant for the ventilatory phenotype. However, an estimate of the individual risk for developing ventilatory symptoms remains unknown, and whether individuals with a +5 alanine somatic mosaic are also at risk to develop symptoms remains also unknown. Interestingly, significantly reduced PHOX2B transactivation potential on dopamine β hydroxylase (DβH) was previously shown, and correlated with the length of the polyalanine tract for all alanine expansions, except for the +5 alanine expansion. Indeed, the transactivation potential was not significantly reduced in the latter. However, spontaneous formation of oligomers, suggestive of misfolding mutant proteins, may occur for the +5 alanine expansion (20), even though protein misfolding leads to cytoplasmic protein aggregation only from +9 alanine or greater expansions in our previous experiments. These cytoplasmic aggregates recruit heat-shock proteins (HSPs), which may counteract aggregate formation, and significantly rescue DβH transactivation, as observed after treatment with a drug up-regulating HSPs (20, 21). HSPs show dual involvement with inherited unstable signal transducers and with cellular response to stress, therefore coupling expression of a morphogenetic variant to the environment. Indeed, in HSP90 mutant flies, phenotypes genetically determined and previously kept at a cryptic state arose as a threshold character (22). Thus, we can speculate that the cellular chaperone system may be able to efficiently cope with the shortest alanine expansion proteins, and depending on cellular stress conditions or individual chaperone activity, this HSP-mediated chaperoning may ultimately result in either congenital (inefficient chaperoning) or late-onset CHS (effective chaperoning). Importantly, phox2b remains highly expressed in neurons of the hindbrain involved in the chemical drive and reflex regulation of the respiratory rhythm generator in adult rat brain (23). In agreement with this theoretical framework, when longer alanine expansions are present, the chaperone systems may be insufficient to cope with misfolded and aggregated proteins, and therefore lead not only to C-CHS but to broader dysfunction of the autonomic nervous system. According to this working hypothesis, germinal alanine expansions from +6 alanines are fully penetrant for the congenital ventilatory phenotype. Here, we show that all asymptomatic carriers of a PHOX2B expansion above +5 alanines are somatic mosaics for the mutation. However, the S194f case raises the question of the risk for all PHOX2B mutation carriers, even if the mutation is somatic mosaic, to develop alveolar hypoventilation. Individual susceptibility is difficult to predict, and could vary principally according to the length of the alanine expansion, the rate of mosaicism in relevant tissues, and age, with environmental factors and modifier gene(s) also potentially involved. PHOX2B frameshift mutations can be divided into two groups according to their translation frame. Interestingly, the +1 out-of-frame translation that leads to an elongated PHOX2B protein predisposes to a severe, fully penetrant phenotype. Indeed, patients reported to date with such germinal mutations present C-CHS, HSCR in the vast majority of the cases, and a high risk for tumor formation in about 50% of cases (12–14, 24). Conversely, the +2 out-of-frame translation that leads to truncated PHOX2B proteins is not fully penetrant for the CHS phenotype. Indeed, such patients have been described with isolated C-CHS, isolated HSCR, Haddad syndrome (C-CHS + HSCR), or TSNS with or without HSCR, and even in asymptomatic carriers (12–14, 25, 26). Accordingly, the frameshift mutation reported here belongs to this latter subgroup. Consistent with these observations, in vitro studies have shown a drastic decrease in the DβH transactivation potential for +1 frameshift and a partial transactivation impairment for the +2 frameshift mutant c.618delC (p.Ser201AlafsX102) (20, 27). Similarly, PHOX2B mutations occurring in the homeodomain of the protein have been reported previously. In two cases, mutation carriers presented C-CHS–HSCR and C-CHS–HSCR-NB for Q143R and R141Q mutations, respectively (12, 14). Two other homeodomain mutations (R100L and R141G) have been reported in familial predisposition to neuroblastoma (NB) with incompletely penetrant HSCR and no ventilatory phenotype (28). In summary, germinal alanine expansion mutations of +6 alanines and above, as well as +1 reading frame mutations lead to a ventilatory phenotype that is both congenital and fully penetrant, whereas germline +5 alanine expansions are not fully penetrant for the ventilatory phenotype and do not lead to either HSCR or to tumor formation. Somatic mosaicism for alanine expansions above +5 may lead in some cases to LO-CHS. Whether this could also be the case for +5 alanine expansion in mosaic remains to be determined. Finally, germinal mutations of the homeodomain and +2 reading frame mutations lead to a wide phenotypic variability. Our group previously showed that a hypomorphic allele of the RET proto-oncogene acts as a modifier gene for the HSCR predisposition in patients with C-CHS (29). On the basis of such an analogy, it is possible that modifier genes may be operational and affect the onset time of the ventilatory phenotype for some PHOX2B mutations. Genes involved in the PHOX2B–RET developmental pathways and genes involved in the chaperone pathway could be therefore regarded as candidate genes (20–22), and environmental factors such as upper airway infections would also likely be involved. Taken together, their interactions could explain the intrafamilial variability observed for some PHOX2B mutations. Finally, the nature of the PHOX2B mutation obviously plays a role. A raised frequency of SIDS in relatives of C-CHS cases has long been debated (30). Families in whom a +5 alanine PHOX2B mutation segregates may confer increased risk for SIDS or apparent life-threatening episodes in relatives. Of note, patient O266, who was found to carry a +5 alanine expansion and who presented LO-CHS, had a brother who died of SIDS. Unfortunately, neither the parents nor the sibling could be tested. Finally, some patients with LO-CHS suffer from broader dysautonomic symptoms, such as ocular anomalies (i.e., strabismus and pupillary anomalies; this report), and reduced heart rate variability as recently reported (31). Our findings further suggest a differential sensitivity for neurons within the ventilatory pathways as compared with those within the enteric nervous system (absent in Hirschsprung disease). Indeed, all types of PHOX2B mutations, even in a mosaic state, can lead to ventilatory symptoms. However, some mutations are more prone to ventilatory phenotype than others (i.e., alanine expansions and +1 frameshift mutations). These observations argue for a gain-of-function mechanism. Consequently, when a PHOX2B mutation is identified in a C-CHS proband, parents should be tested. Those parents who are found to be mutation carriers, either germinal or somatic, should be periodically evaluated, and counseled on the increased risk potentially brought about by situations such as respiratory infections or anesthesia. Drugs known to depress ventilation should also be preferentially avoided, or given with extreme caution. Finally, PHOX2B gene mutations should be systematically examined in any adult with unexplained central hypoventilation.
The authors thank the patients and their families for their active participation in the study and their treating physicians for referring them.
Supported by the Agence Nationale pour la Recherche. D.T. is funded by the Fondation pour la Recherche Médicale. Originally Published in Press as DOI: 10.1164/rccm.200707-1079OC on December 13, 2007 Conflict of Interest Statement: D.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. L.d.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.S. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. D.G. has received $12,000 in 2006 and $6,000 in 2007 for conferences sponsored by Merck Co. H.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. P.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.M. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. S.L. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. C.G. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.A. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form July 22, 2007; accepted in final form December 13, 2007
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
90% of the cases). Frameshift mutations or missense mutations within the homeodomain of the protein are found in a minority of cases (
