Published ahead of print on April 22, 2004, doi:10.1164/rccm.200402-245PP
© 2004 American Thoracic Society
Genetics of Congenital Central Hypoventilation SyndromeLessons from a Seemingly Orphan DiseaseDepartments of Pediatrics and of Neurology and Biochemistry, Rush Children's Hospital at Rush University Medical Center, Chicago, Illinois Correspondence and requests for reprints should be addressed to Debra E. Weese-Mayer, M.D., Pediatric Respiratory Medicine, Rush Children's Hospital, 1653 West Congress Parkway, Chicago, IL 60612. E-mail: Debra_E_Weese-Mayer{at}rsh.net The initial response to a publication on a pediatric "orphan disease" is not necessarily enthusiasm, especially if the reader is an adult intensivist or pulmonologist. Although considered insignificant because of their rarity, orphan diseases can serve as models for other, more global phenomena. For example, idiopathic congenital central hypoventilation syndrome (CCHS), although first described in 1970 (1), was considered exclusively as a rare disorder of respiratory control, with nearly all publications in case report or small series format for the next 20-plus years. Although many early reports described symptoms compatible with autonomic nervous system (ANS) dysregulation, and reports on familial patterns appeared, it took another decade for the familial nature of CCHS and the associated ANS dysregulation to be identified. The American Thoracic Society increased the visibility of CCHS with the 1999 ATS statement entitled "Idiopathic Congenital Central Hypoventilation Syndrome: Diagnosis and Management" (2). The specific objectives were to (1) improve knowledge regarding CCHS to minimize delays in diagnosis; (2) state the components of a thorough diagnostic evaluation and for follow-up; (3) familiarize practitioners with diagnostic, treatment, and home health care options, and long-term outcomes; (4) emphasize the composite of ANS dysfunction beyond the narrow control of breathing deficit considered the hallmark of CCHS; and (5) provide evidence of a genetic component. Subsequent to the ATS statement we have documented an increase in the number of cases diagnosed, with earlier consideration of the disease. Nearly all subsequent publications have appeared in the genetics literature, leaving the readership of the American Journal of Respiratory and Critical Care Medicine potentially uninformed regarding the remarkable progress in determining the genetic basis of CCHS as well as the inheritance pattern. This Pulmonary Perspective has been prepared for three reasons: (1) to demonstrate the lesson that a seemingly straightforward orphan disease with disordered respiratory control is actually a more global phenomenon of ANS dysregulation; (2) to bring the remarkable genetic progress in CCHS, culminating with the identification of a defining gene, the paired-like homeobox gene (PHOX)2b, to the adult and pediatric pulmonologist/intensivist readership of the Journal; and (3) to demonstrate that this unique study population holds great promise for furthering our understanding of the ANS, and of ANS dysfunction. Some of the results of these studies have been previously reported in the form of an abstract (3). CLINICAL DESCRIPTION OF CCHS Current records indicate that about 300 children worldwide have CCHS, although this number is likely an underestimate. CCHS is diagnosed in the absence of primary neuromuscular, lung, or cardiac disease, or of an identifiable brainstem lesion. It is characterized by generally adequate ventilation while the patient is awake but hypoventilation with typically normal respiratory rates and diminished tidal volume during sleep (2). On occasion, these patients will demonstrate apneic pauses after discontinuation of mechanical ventilation and before initiation of spontaneous breathing. More severely affected children hypoventilate both awake and asleep. While asleep, children with CCHS experience progressive hypercapnia and hypoxemia (2), although their ventilation is "more normal" in rapid eye movement sleep than in nonrapid eye movement sleep (4). They have absent or negligible ventilatory sensitivity to hypercarbia and absent or variable ventilatory sensitivity to hypoxemia during sleep (2). They lack an arousal response to the endogenous challenges of isolated hypercarbia, hypoxemia, and to the combined stimulus of hypercarbia and hypoxemia (2). Awake ventilatory responsiveness to and perception of hypercarbia and hypoxemia are generally absent (2), even when awake minute ventilation is adequate. Children with CCHS lack a perception of dyspnea but maintain conscious control of breathing (5). During exercise these children may be at risk for hypercarbia and hypoxemia, although the degree of exercise and the severity of the CCHS likely impact on the individual response (5–7). These patients demonstrate sustained abnormalities in control of breathing and continue to require long-term ventilatory support, with several patients entering early adulthood. Conditions associated with CCHS include Hirschsprung disease and tumors of neural crest origin: ganglioneuroma, neuroblastoma, and ganglioneuroblastoma. DATA TO SUPPORT ANS DYSREGULATION IN CCHS Children with CCHS have a complex phenotype consistent with imbalance of the ANS, yet it has taken several years for acceptance that CCHS is a more global phenomenon rather than strictly an abnormality of control of breathing. Many of the early case reports describe decreased heart rate variability, diminished pupillary light response, feeding difficulty with esophageal dysmotility in infancy, breath-holding spells, poor temperature regulation with decreased basal body temperature, and sporadic profuse sweating episodes with cool extremities. Table 1 indicates symptoms of ANS dysfunction among 56 children with CCHS and their age-, sex-, and ethnicity-matched but unrelated control subjects (8). Yet only in the past year has methodologic ANS testing progressed beyond heart rate variability assessment, with two new important physiologic studies describing esophageal dysmotility (9) and blood pressure fluctuation (10). Development of proper laboratories to assess ANS function in children is essential to clearly characterize the phenotype of ANS imbalance in CCHS and other potentially related diseases.
DATA TO SUPPORT A GENETIC BASIS FOR CCHS Familial recurrence of CCHS includes one report each of affected monozygotic female twins (11), sisters (12), male–female sibs (13), and male–female half-sibs (14). Segregation analysis of a quantitative ANS dysfunction trait in CCHS families revealed that the best-fitting model for ANS dysfunction was codominant Mendelian inheritance of a major gene (15). Report of a child with CCHS born to a woman who had neuroblastoma as an infant (16) supports evidence of a transmitted genetic component in the phenotypic spectrum of ANS dysfunction and CCHS. Description of five women, diagnosed with CCHS in their own childhoods, who gave birth to four infants with CCHS (17, 18) provided additional evidence of the genetic basis for CCHS. GENETIC SCREENING OF GENES PERTINENT TO ANS DYSREGULATION IN CCHS Because Hirschsprung disease occurs in about 20% of individuals with CCHS, the initial molecular analyses in CCHS were directed at genes known to be related to Hirschsprung disease, including receptor tyrosine kinase (19, 20), endothelin signaling pathway genes (21–24), and glial-derived neurotrophic factor (25, 26). Subsequently, screening was expanded to genes known to be involved in neural crest cell migration and in ANS development in the brain during embryogenesis. Twenty patients have been reported with unique protein-altering mutations in other genes: eight in tyrosine kinase (27–31); one in glial-derived neurotrophic factor (27); one in endothelin-3 (32); one in brain-derived neurotrophic factor (33); five in human achaete-scute homolog-1 (31, 34); one in PHOX2a (31); one in GFRA1 (31); one in bone morphogenic protein-2 (35); and one in endothelin-converting enzyme-1 (35). Because many of these mutations occur in family members who are not affected with CCHS, it is not clear whether these variants are only partially penetrant, represent risk factors for CCHS, or are entirely unrelated to the phenotype. IDENTIFICATION OF PHOX2b AS A DISEASE-DEFINING GENE FOR CCHS Amiel and coworkers reported heterozygous expansion mutations in a polyalanine tract within exon 3 of PHOX2b in 18 of 29 children (62%) with CCHS in France (36). We subsequently reported the exon 3 polyalanine expansion mutation in 65 of 67 children (97%) with CCHS in a large, predominantly American cohort (Figure 1) , utilizing a simple and accurate assay for sizing the repeat sequence associated with the polyalanine tract expansion (35). One further report (31) identified the polyalanine repeat expansion mutation in 4 of 10 Japanese children (40%) with CCHS. Each of these children with the mutation has 1 normal allele with 20 repeats of the polyalanine sequence and 1 allele with extra repeats. The range for number of repeats in the expansion is 25–29 (36), 25–30 (31), and 25–33 (35). The polyalanine expansion mutation was not found in any of 242 control subjects from the above-cited publications. Deletion variants with only 14 or 15 repeats in the polyalanine repeat tract have been reported in 3 children with CCHS (31, 35), but are also found in about 3% of normal control subjects and CCHS parents. The PHOX2b polyalanine repeat expansion mutation was also observed in 9 of 16 patients with receptor tyrosine kinase, glial cell-derived neurotrophic factor, brain-derived neurotrophic factor, human achaete-scute homolog-1, and GFRA1 mutations. Collectively, these data strongly suggest that PHOX2b is the disease-defining gene in CCHS.
Further evidence that PHOX2b is the primary CCHS gene comes from reports of children with no polyalanine repeat expansion but other PHOX2b mutations. In the French cohort, 2 of 11 children had frameshift mutations (37 nucleotide deletion and a single base pair insertion) in exon 3 of PHOX2b. In our cohort, one of two children had a protein-truncating nonsense mutation in exon 3. In the Japanese cohort, one of six children had a frameshift mutation (single base pair insertion) in exon 3. INHERITANCE OF PHOX2b MUTATIONS AND CCHS Parents of children with CCHS and children born to probands with CCHS have also been studied. Amiel and coworkers (36) and Sasaki and coworkers (31) studied eight and two sets of parents, respectively, and then concluded that the PHOX2b mutation occurs de novo. We studied 54 families (including 97 unaffected parents), and identified 4 parents (about 10%) with mosaicism for an expansion mutation identical to that seen in their offspring with CCHS, suggesting that not all mutations in affected probands with unaffected parents are de novo (Figure 1) (35). We also studied four women with CCHS and their offspring, identifying three mother–infant pairs with an identical PHOX2b mutation. Taken together, these data demonstrate autosomal dominant inheritance of the heterozygous mutation (Figure 1) (35). Hence, there would be up to a 50% risk of CCHS for future offspring of mosaic parents of an existing child with CCHS . For offspring born to probands with CCHS, the risk of CCHS would be 50%. POSSIBLE EXPLANATION FOR DISPARITIES IN INCIDENCE OF PHOX2b MUTATIONS IN CCHS We identified the PHOX2b polyalanine expansion mutation in 97% of cases and any PHOX2b mutation in 98.5% of 67 CCHS cases (35), as opposed to 62 and 69% (36) and 40 and 50% (31) of 29 and 10 cases, respectively. Because 41% of the French cases and 30% of the Japanese cases had Hirschsprung disease and/or a tumor of neural crest origin versus 24% of our cases at Rush University Medical Center (Chicago, IL) (35), some of the subjects without the PHOX2b mutation may represent more of an anatomic (versus a physiologic and anatomic) neurocristopathy. This could result in differences in the clinical diagnostic criteria for CCHS among centers. Autonomic dysfunction has been described in a cohort of children with Hirschsprung disease (but no CCHS) (37), and therefore the observation that six of nine French subjects and three of five Japanese subjects without any PHOX2b mutation had Hirschsprung disease, or severe constipation, raises the possibility that these patients present phenotypes similar to that of children with CCHS (i.e., Hirschsprung disease with autonomic dysfunction [37] who also have alveolar hypoventilation). The report of Garcia-Barceló and coworkers (38) reporting a PHOX2b polymorphism associated with Hirschsprung disease (unrelated to CCHS) strengthens this possibility. POTENTIAL MECHANISMS AND IMPACT OF THE PHOX2b MUTATION(S) ON NEURODEVELOPMENT OF THE ANS IN CCHS PHOX2b is mapped to chromosome 4p12, and encodes a highly conserved homeobox domain transcription factor (314 amino acids), with two short and stable polyalanine repeats of 9 and 20 residues (36). It has an early embryologic action as a transcriptional activator involved in promoting pan-neuronal differentiation including upregulation of proneural gene and mammalian achaete-scute homolog-1 expression, and expression of motoneural differentiation (39). Through an alternative pathway, PHOX2b has a separate role wherein it represses expression of inhibitors of neurogenesis (39). Finally, PHOX2b is required to express tyrosine hydroxylase, dopamine ß-hydroxylase (40), and receptor tyrosine kinase and to maintain mammalian achaete-scute homolog-1, thereby indicating that PHOX2b regulates the noradrenergic phenotype in vertebrate neural cells (41). Extensive studies by Pattyn and coworkers (41, 42) indicate an early expression pattern of PHOX2b in rhombencephalon, suggesting a link to early patterning events with later neurogenesis in the hindbrain. In the mouse, Phox2b is expressed in the neonatal CNS, specifically in the area postrema, nucleus tractus solitarius, dorsal motor nucleus of the vagus, nucleus ambiguus, ventral surface of the medulla, locus coeruleus (until embryonic day 11.5), and cranial nerves III, IV, VII, IX, and X. Until midgestation in the mouse, Phox2b is expressed in the Vth cranial nerve. In the mouse peripheral nervous system, Phox2b is expressed in the distal VIIth, IXth, and Xth cranial sensory ganglia from embryonic day 9.5 and in all ANS ganglia as early as formed, until at least midgestation. By embryonic Day 9–9.5, Phox2b protein is detected in enteric neuroblasts invading the foregut mesenchyme. It is expressed in the esophagus, small intestine, and large intestine (41, 42), and in all undifferentiated neural crest-derived cells in the gut with a rostrocaudal gradient (43). In the Phox2b knockout, the gut is devoid of enteric neurons, the neural crest-derived cells found in the foregut at embryonic day 10.5 do not survive or migrate further (43), and the neural crest-derived carotid body and the visceral sensory ganglia degenerate while the nucleus of the solitary tract never forms (44). In addition, the Phox2b heterozygous mouse mutants have an altered response to hypoxia and hypercapnia, and decreased tyrosine hydroxylase expression in petrosal chemosensory neurons. These data establish Phox2b as a key transcription factor in the formation of autonomic reflex pathways and suggest a mechanism for CCHS (44) accounting for the seeming imbalance in the sympathetic and parasympathetic nervous systems, and relative dysfunction in the enteric nervous system of children with CCHS. Polyalanine expansion mutations have been described as a cause of disease in a number of homeodomain- and non–homeodomain-containing transcription factors (45). Polyalanine repeat tract expansion in the Aristaless-related homeobox gene results in X-linked mental retardation and epilepsy due to presumed failure of specification and/or migration of GABA-ergic neurons (46–48). Given the expression of PHOX2b in central autonomic structures and peripheral neural crest derivatives and the wide range of ANS dysfunction seen in CCHS, it seems plausible that a similar mechanism of abnormal differentiation and/or migration of central and peripheral noradrenergic sympathetic, parasympathetic, and enteric neurons could result from mutations in PHOX2b. The identification of mutations in PHOX2b that cause a CCHS phenotype and produce a truncated protein (35) or a highly disrupted protein due to out-of-frame intragenic deletion (36) suggests that the polyalanine expansion mutation results in CCHS through a loss-of-function mechanism. However, no mutations that disrupt the homeodomain have thus far been identified in CCHS, raising the possibility that if these abnormal proteins are stable, they may still have some function. Indeed, polyalanine tracts are thought to act as spacers or protein-binding elements. Therefore, expansions of these tracts in a mutant protein that can still bind DNA could prevent normal protein interactions (loss of function) or allow aberrant interactions (gain of function) while blocking function of the normal protein encoded by the nonmutated gene (dominant negative). This would explain the dominant inheritance pattern in a loss-of-function mutant. In fact, the description of PHOX2b homeodomain mutations in two families with tumors of neural crest origin, but not CCHS, suggests that a loss-of-function mechanism may be sufficient to predispose to such tumors (49). In contrast, the full CCHS phenotype may be dependent on mutations, distal to the homeodomain region, that are more likely to produce a dominant-negative or gain-of-function effect. Further, in these models, longer repeat tracts might be expected to produce a more pronounced molecular disturbance, resulting in increasing severity and number of clinical symptoms with increased length of repeat tract. PRACTICAL APPLICATIONS OF IDENTIFICATION OF PHOX2b MUTATIONS IN CCHS There is now a simple DNA test for diagnosis of CCHS, which can be run on DNA from blood samples or other tissues. If the test is negative and the physician is confident that the child has the phenotype for CCHS, then sequencing of the PHOX2b gene should be performed. Because of the autosomal dominant inheritance pattern, it would be advisable to perform the test on parents of CCHS probands and on probands with CCHS who are pregnant. Prenatal testing for CCHS can be done on cultured chorionic villus sampled tissue or amniocytes if the PHOX2b mutation in the family is known. The PHOX2b test will now allow for identification of a uniform population for clinical studies of CCHS. Such studies should lead to further clarification of the relationship of repeat number for the polyalanine expansion to phenotype, as well as the relation of other mutations in PHOX2b to phenotype (e.g., if certain types of mutations are more likely to result in tumors of neural crest origin and Hirschsprung disease). GENOTYPE/PHENOTYPE CORRELATION IN CCHS In contrast to the study by Amiel and coworkers (36), we identified an association between PHOX2b polyalanine repeat expansion mutation length and number of ANS dysfunction symptoms (p = 0.021) in CCHS probands (Figure 2) . Further, among 23 patients with CCHS and available Holter monitor data, the mean R–R interval (2.30 ± 1) in 11 patients with 26 polyalanine repeats was significantly different from the mean R–R interval (3.78 ± 1) in 12 patients with 27 repeats (p = 0.001), suggesting that an increased number of repeats of the polyalanine expansion mutation may be associated with longer sinus pauses in patients with CCHS (3). Amiel and coworkers (36) likely did not identify relationships between CCHS/ANS dysfunction symptoms and genotype because of the smaller sample size studied.
FUTURE DIRECTIONS With phenotype–genotype characterization in future studies, it may be possible to further characterize how this mutation in a gene pertinent to early embryologic development of the ANS affects physiologic outcome in children with CCHS. Continuing genetic studies in children with alveolar hypoventilation and ANS dysfunction, but no PHOX2b mutation, may allow identification of additional candidate genes in signaling pathways and molecular events leading to neural cell differentiation and development that further refine our understanding of ANS development. Many children with CCHS are now young adults, making clarification of these phenotype–genotype relationships important for long-term clinical management. FOOTNOTES Conflict of Interest Statement: D.E.W.-M. is named on a patent application for the PHOX2b assay described in Reference 35 and, to date, has not received any financial benefit from this application; E.M.B.-K. is named on a patent application for the PHOX2b assay described in Reference 35 and, to date, has not received any financial benefit from this application. Received in original form February 25, 2004; accepted in final form April 20, 2004 REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
