Hirschsprung's disease (OMIM 142623) is a complex congenital disorder characterised by the absence of ganglion cells of the plexus myentericus and plexus submucosus in the variable lengths of the digestive tract.1,2 Aganglionosis is attributed to a defect of the enteric nervous system, in which ganglion cells fail to innervate the lower gastrointestinal tract during embryonic development, resulting in failure to pass meconium, chronic severe constipation and colonic distention in the neonatal period.3 The receptor tyrosine kinase gene RET, which is expressed in neural crest cells during enteric neurogenesis and is required for normal development of the enteric nervous system, is the major susceptibility gene for Hirschsprung's disease.4 There is growing evidence indicating that functional single nucleotide polymorphisms (SNPs) of RET could act as low susceptibility factors for Hirschsprung's disease.5,6 In addition, PHOX2B encodes a transcription factor that is involved in the development of the noradrenergic nervous system and that plays an important role in the regulation of RET transcription.7,8 Our previous data showed that genotypes comprising allele A of the IVS2+100 A>G SNP of PHOX2B were associated with an increased risk of Hirschsprung's disease.9 In view of the role of RET and PHOX2B in the development of the enteric nervous system, we hypothesised that RET and PHOX2B polymorphisms are likely to interact and have a joint effect in conferring susceptibility to Hirschsprung's disease.
This case–control study consisted of 256 ethnic Chinese patients histologically diagnosed with sporadic Hirschsprung's disease, including 13 patients with total colonic aganglionosis, 28 with long‐segment aganglionosis and 215 with short‐segment aganglionosis. Controls were 242 unselected, unrelated, ethnic Chinese subjects. At recruitment, informed consent was obtained from each subject. This study was approved by the institutional review board of the University of Hong Kong. Genotypes for RET promoter polymorphisms (5G>A and 1A>C) and PHOX2B were analysed using polymerase chain reaction and direct sequencing, as described previously.6,9
We observed an increased risk of Hirschsprung's disease in homozygous genotypes of the disease‐associated RET alleles 5AA or 1CC when compared with 5GA and 5GG (OR=7.78, 95% CI 5.21 to 11.70), or 1AC and 1AA (OR=6.08, 95% CI 4.01 to 9.12) genotypes. The same risk was seen with the PHOX2B IVS2+100 AA genotype compared with the GG or GA genotypes (OR=1.75, 95% CI 1.17 to 2.58). We then investigated the potential interaction between the RET and PHOX2B genotypes in the risk of Hirschsprung's disease using the additive interaction models (table 11).10 Since the 5 and 1 RET SNPs were in almost complete linkage disequilibrium in our population (D'=0.986, p<0.001), only the former was selected for further analysis. Among subjects carrying at least one RET 5G allele or at least one PHOX2B G allele, the OR for Hirschsprung's disease was 1.56 (95% CI 0.82 to 2.75) or 7.98 (95% CI 3.64 to 17.60). The OR increased, however, to 11.72 (95% CI 6.35 to 21.55) among subjects carrying both the RET 5AA and PHOX2B AA genotypes (p<0.001, test for homogeneity). The OR value of 11.72 is larger than the values of the RET 5AA and the PHOX2B AA genotypes independently (7.78+1.75−1=8.53), which indicates a more than additive interaction between the RET and the PHOX2B SNPs on risk of developing Hirschsprung's disease according to the statistical model.10
To our knowledge, this is the first study to demonstrate that the interaction between RET and PHOX2B polymorphisms has a substantial impact on risk of Hirschsprung's disease. This recognised multifactorial genetic disorder requires the interaction of several unlinked genes to produce the phenotype. Both RET and PHOX2B have important roles in the development of the enteric nervous system and PHOX2B is involved in the transcriptional regulation of RET in cell lines originated from neural crest. It is therefore biologically plausible that a joint effect of RET and PHOX2 SNPs affecct the risk of Hirschsprung's disease. However, the essential mechanisms behind our finding need to be investigated.