A search for Hirschsprung disease in OMIM under the symbol HSCR reveals 12 entries covering both phenotypes and the specific genes involved. The phenotypes refer to HSCR1 (MIM 142623), largely comprising RET gene mutations; HSCR2 (MIM 600155), comprising EDNRB gene mutations; HSCR3 (MIM 600156), comprising a modifier on chromosome 21; and, congenital failure of autonomic control (MIM 209880). GenBank accession numbers: RET—AL022344, X12949, X15262, AJ243297; GDNF—L19063; EDNRB—D90402, AL139002; EDN3—J05081, AL035250; SOX10—AJ001183, AL031587; NTN—U78110; ECE-1—D43698, AL031005.
Hirschsprung disease (HSCR), or congenital aganglionosis, is defined by the absence of intramural ganglion cells in the myenteric and submucocal plexuses of the gastrointestinal (GI) tract. The disorder is classified into long-segment (L-HSCR: aganglionosis of the splenic flexure and beyond), colon-segment (aganglionosis of the descending colon), and short-segment (aganglionosis of the sigmoid colon and more distal regions) forms; a minority of patients have total colonic aganglionosis (TCA). The diagnosis of HSCR occurs in the neonatal period at a median age of 7.5 days. Clinical symptoms in neonates are variable and commonly include severe constipation, abdominal distension, and failure to pass meconium. Generally, diagnosis involves suction rectal biopsy, but full-thickness biopsy is necessary for differential diagnosis. Contemporary surgical treatment, placement of the bowel with normal peristalsis at the anus to eliminate the tonic contraction of the internal sphincter, has led to excellent prognosis and a normal life span in >75 percent of patients. Nevertheless, enterocolitis and chronic constipation remain as long-term complications in many patients.
HSCR is a neurocristopathy or a disorder of neural crest (NC) cells. The NC is a transient and multipotent embryonic structure that gives rise to neuronal, endocrine and paraendocrine, craniofacial and conotruncal heart, and pigmentary tissues. In particular, the enteric nervous system (ENS) is formed from the craniocaudal migration of vagal-derived NC cells during weeks 5 to 12 of gestation. The embryologic development of the ENS is genetically programmed and known to depend on several proteins which determine and facilitate the orderly migration, proliferation, differentiation, and survival of NC cells in the walls of the GI tract to form the myenteric and submucocal plexuses. The details of ENS development explain phenotypic features of HSCR as well as its association with other syndromes involving the NC.
The population incidence is ≈1/5000 live births with sex-ratio 3:1 male:female; however, there is considerable population variation in both incidence and sex-ratio and other patient characteristics. Among all cases, 18 percent have L-HSCR and 7 percent have TCA. In 70 percent of patients HSCR occurs as an isolated trait, 12 percent have a recognized chromosomal abnormality and 18 percent multiple congenital anomalies. The most common (>90 percent) chromosomal abnormality is trisomy 21, but multiple deletions of segments of chromosomes 2, 10, 13, and 17 have been noted. The congenital anomalies, beyond those associated with trisomy 21, common to HSCR include atresia or stenosis of the GI tract, polydactyly, cleft palate, cardiac septal defects, and craniofacial anomalies, defects that are explainable from NC biology. A critical finding is the higher frequency of multiple anomalies in familial (39 percent) than in isolated (21 percent) cases, suggesting underdiagnosis of specific features in the HSCR patient and warranting careful phenotypic assessment of each patient irrespective of family history. No known environmental factor has been associated with HSCR, which largely appears to be a genetic defect in NC-derived tissues.
Family history studies have established the recurrence risks of HSCR to be ≈3 percent and ≈17 percent for S-HSCR and L-HSCR, respectively, corresponding to heritability of ≈100 percent for either form. The increasing survival of surgically repaired patients have led to the identification of numerous multiplex families and parent-offspring transmissions. Genetic modeling has shown L-HSCR and colonic-HSCR to be inherited as rare autosomal dominant traits with low penetrance and with a substantial fraction of sporadic cases; S-HSCR is inherited as either a recessive trait with very low penetrance and minimal sporadic cases or has multifactorial inheritance. The segregation of HSCR with features of Waardenburg syndrome (Shah-Waardenburg syndrome) also shows dominant inheritance with lower penetrance of aganglionosis than the Waardenburg features. In all analyses, the penetrance of the putative gene mutants is greater in males than in females.
Specific genes involved in HSCR have been identified by linkage studies, positional cloning, and mutation analysis of candidate genes. In humans, five genes (the receptor tyrosine kinase [RET], its ligand GDNF , the G-protein-coupled receptor EDNRB , its ligand EDN3 , and the transcriptional regulator SOX family transcription regulator gene type 10 (SOX10) ) have been shown to harbor multiple mutations in HSCR. In addition, the alternative RET ligand NTN and the endothelin-processing enzyme ECE1 have single mutations in HSCR. Rare mutations in EDN3, RET, or GDNF are also associated with congenital central hypoventilation syndrome (CCHS). Disease gene mapping in families segregating mendelian forms of HSCR have also revealed a role for the SMAD interaction protein 1 (ZFHX1B) in the genesis of rare syndromic forms of HSCR associated with midline abnormalities. The large majority of mutations in HSCR have been recognized in RET, which is also the gene mutant in multiple endocrine neoplasia type 2 (MEN2). Interestingly, HSCR RET mutations are loss-of-function alleles, while MEN2 RET variants are activating mutations. The genetic features of all mutations detected show that mutations in the RET pathway are haploinsufficient and HSCR occurs in mutation heterozygotes; mutations in the EDNRB pathway are hypomorphic and pleitropic because HSCR occurs in mutation heterozygotes, but Shah-Waardenburg syndrome occurs in mutation homozygotes. Mutations in SOX10 are haploinsufficient and pleitropic because Shah-Waardenburg syndrome occurs in mutation heterozygotes. All mutations identified show greater penetrance in males than in females, are detected with greater frequency in L- than in S-HSCR, and in familial than in sporadic cases. However, mutations are seldom identified in more than 30 percent of HSCR cases. These studies establish the roles of RET and EDNRB signaling and SOX10 regulation as necessary for normal ENS development.
Mouse models of aganglionosis and megacolon have been critical to the identification of human HSCR genes and in understanding ENS development. Homozygotes for null mutations in Ret, Gdnf, Ednrb, and Edn3, and heterozygotes for a null Sox10 mutation, demonstrate aganglionosis in the mouse. Moreover, homozygotes for null mutations in Ece1, A-raf, Ncx, Hoxa4, and Dlx2 also show aganglionosis in the mouse. These models emphasize the critical genes that determine ENS innervation but depart from the human phenotype in that none of them display reduced sex-dependent penetrance and none, except Sox10, show aganglionosis in heterozygotes. However, two-locus (Ret/Ednrb) mouse strains recently have been reported to recapitulate the sex bias and incomplete penetrance observed among HSCR patients.
Genotype-phenotype correlation in either L-HSCR or S-HSCR is very poor and is compounded by the reduced sex-dependent penetrance. Rare families in which the existence of multiple known mutations can explain trait segregation has led to the search for common modifying genes in HSCR. Recent studies show the segregation of a modifier on human chromosome 9q31 in L-HSCR families known to harbor a RET mutation; specifically, segregation of the 9q31 locus was restricted to families having atypical or noncoding mutations. In addition, in S-HSCR the segregation of three loci at 3p21, 10q11, and 19q12 appear to be both necessary and sufficient for explaining HSCR segregation and incidence. The susceptibility factor at 10q11 is RET, but typical coding sequence mutations could be identified in only 40 percent of linked families. Furthermore, a recent genome-wide association study in an Old Order Mennonite kindred uncovered interaction between RET and a known hypomorphic mutation in EDNRB as a mechanism in HSCR. The majority of HSCR is oligogenic, involves RET in almost all cases, but involves atypical noncoding variants.
HSCR is frequently found in association with many other neurocristopathy syndromes, such as Waardenburg and other pigmentary syndromes, CCHS, MEN2, and with neural tube defects. Additional significant associations include the Goldberg-Shprintzen syndrome and numerous others involving distal limb anomalies. These associations can be explained from our current knowledge of NC cell development and suggests an intrinsic defect, inherited or sporadic, of NC cells. HSCR is occasionally observed in patients with Bardet-Biedl syndrome, Smith-Lemli-Opitz syndrome types 1 and 2, and cartilage hair hypoplasia, which, being monogenic, suggests that some genes in common may predispose to both traits. The majority of the other common HSCR associations are chromosomal abnormalities, which suggests the action of dosage sensitive genes lying within the affected genomic segment. These syndromic associations reveal common development pathways affecting the NC and its derivatives and suggest specific candidate genes for HSCR.
The significant clinical and genetic knowledge in HSCR suggests that genetic counseling of patients and their families should utilize the known variation in recurrence risk by proband gender, consultand gender, segment length involved, familiality and association with other disorders. Current data suggests that counseling for Shah-Waardenburg syndrome may follow standards for monogenic disorders, while patients with RET exon 10 and 11 mutations (at residues also mutant in MEN2) warrant further testing to ascertain predisposition to neuroendocrine tumors. The emerging evidence for the centrality of RET, in perhaps all HSCR cases, and oligogenic inheritance may allow family specific risks to be estimated in the near future.