Introduction

Hirschsprung disease (HSCR) affects approximately 1:5000 live births (the highest incidence is observed in Asian population, 2.8:10000 live births) and is characterized by the absence of parasympathetic ganglion cells (aganglionosis) in the hindgut, leading to tonic contraction of the affected segment, intestinal obstruction and massive distension of the proximal bowel [1,2]. HSCR is considered as a neurocristopathy since it is caused by the premature arrest of the craniocaudal migration of vagal neural crest cells in terminal regions of the gut between the 5^th and 12^th week of gestation to form the enteric nervous system (ENS) [1]. It can be anatomically categorized into three types, according to the extent of aganglionosis: short-segment HSCR (S-HSCR, 80% of cases) which affects the rectum and a short portion of the colon, long-segment HSCR (L-HSCR, 15% of cases) which affects longer tracts of the colon and total colonic aganglionosis (TCA, 5% of cases) [3].

The genetic complexity observed in HSCR can be attributed to non-Mendelian inheritance in nature, low sex-dependent penetrance, variability in the length of the ananglionic segment and involvement of multiple genetic and environmental factors [4,5]. Variants of at least eleven genes to date have been implicated in Hirschsprung disease including the two major ones, RET (receptor tyrosine kinase) and EDNRB (endothelin receptor type B), whereas these mutations account for only 50% of familial and up to 20% sporadic subjects with HSCR, and cumulatively explain a small proportion of the heritability [5]. Additionally, common variants in the regulatory region of RET were found to be involved in mediating susceptibility to HSCR by conferring an obvious reduction of the RET expression [6]. On the other hand, joint gene-gene effects (e.g. RET and PHOX2B; RET and HOX genes) may also have a substantial impact on the risk of Hirschsprung disease [7,8]. Recently, NRG3 that performs important functions during neural development has been identified as a novel susceptibility gene in a whole exome sequencing study [9]. However, which of the candidate genes encompassed by these susceptibility loci mainly contribute to HSCR susceptibility remains unresolved.

Patched 1 (PTCH1) gene located on chromosome 9q22, 3 is expressed in three major isoforms (1, 1A and 1B), which are upregulated by transcription factor Gli [10]. Since PTCH1 functions as a tumor suppressor, Danaee and colleagues have found that loss of PTCH1 function due to mutation or deletion contributed to the genesis of basal cell carcinoma [11]. The encoded protein, PTCH1, serves as Hedgehog (Hh)-binding receptor, and is supposed to be the key regulator of the Hedgehog (Hh) signaling pathway that has been implicated in mediating proliferation and differentiation of the enteric neural crest cells (ENCCs) [12,13]. Reduced PTCH1 causes overstimulation of Hh which can result in the hyperproliferation of epithelial cells [12]. By applying canonical correlation analysis, Ngan et al. [13] found mutations within PTCH1 conferred higher risk to Hirschsprung disease, and deletion of Ptch1 in mouse ENCCs led to premature gliogenesis and reduction of ENCC progenitors in mutant bowels. In the present study, we aimed to evaluate whether common variants in the PTCH1 gene might contribute to the altered risk of Hirschsprung disease in the Han Chinese population.

Materials and Methods

SNP selection and genotyping

We recruited eight genetic polymorphisms, namely rs357552 and rs10512248, which had been reported by Ngan et al. [13], and six other SNPs (rs357565, rs28485160, rs357564, rs2236405, rs28701981 and rs2236407) from HapMap project database (http://www.hapmap.org) and dbSNP (http://www.ncbi.nlm.nih.gov/SNP/) to cover the region of PTCH1. The eight genetic markers involved in the study spanned ^~ 73.4 kb region of PTCH1, with an average interval of ^~ 9.2 kb: rs357565 and rs28485160 in the untranslated region of exon1, rs357564 in exon2, rs2236405 in exon3, and rs357552, rs28701981, rs2236407, and rs10512248 in the intronic region (Figure 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Distributions of the eight genetic polymorphisms in the genomic region of PTCH1.

Red lines represent the SNPs genotyped; blue lines and arrow indicate all the 24 exons within PTCH1.

https://doi.org/10.1371/journal.pone.0075407.g001

All genetic polymorphisms were genotyped using the MassARRAY iPLEX Gold technology (Sequenom, San Diego, CA). Briefly, PCR and single-base extension primers (SBE) were designed using the Assay Design Suite of Sequenom. The whole process was performed according to the manufacture’s instructions for the multiplex reaction, including the PCR amplification, the shrimp alkaline phosphatase (SAP) treatment, and the primer extension reactions using iPLEX Gold assay (Sequenom, San Diego, CA). Reaction products were then dispensed onto a 384-SpectroCHIP using the MassARRAY Nanodispenser and analyzed on the MassARRAY platform. Mass signals for the different alleles were captured by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) with high accuracy. Typer Version 4.0 (Sequenom, San Diego, CA) was used to process raw data obtained from the assays.

Results

As for the studied genetic markers, Hardy-Weinberg equilibrium tests were performed in cases and controls respectively. Genotype distributions were in Hardy-Weinberg equilibrium for all eight polymorphisms in either cases or controls (P > 0.05). For the 104 subjects with Hirschsprung disease and the 151 normal controls, we found that rs357565 and rs2236405 were significantly associated with Hirschsprung disease (rs357565, allele, P = 0.005, genotype, P = 0.015; rs2236405, allele, P = 0.002, genotype, P = 0.003). Representative mass spectra of the original MassArray reactions (e.g. SNP rs357565) are presented in Figure 2. Allele and genotype distributions of the 8 SNPs are shown in Table 2. The results observed in rs357565 and rs2236405 remained significant after the FDR correction (rs357565, allele, P = 0.018; rs2236405, allele, P = 0.019, genotype, P = 0.028). Of note, the C allele and CC genotype of rs357565 were more common in the case group compared to the control group (C allele, 87.3% versus 77.2%, OR = 0.49, 95% CI 0.30-0.81; CC genotype, 76.5% versus 59.1%), and likewise the A allele and AA genotype of rs2236405 showed significantly higher frequencies in the HSCR subjects than in the normal controls (A allele, 18.4% versus 9.0%, OR = 2.28, 95% CI 1.33-3.93; AA genotype, 6.1% versus 0%). Since there was a difference in the proportions of males and females between the two subject groups, PLINK software was included in the adjustment for age and gender factors, and the significance of the two markers remained after correction (P < 0.05).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Representative mass spectra of the MassArray reactions including the three genotypes (A: A/A; B: C/C; C: A/C) of SNP rs357565.

The alleles and UEP (unextended primer) are indicated with thick horizontal arrows.

https://doi.org/10.1371/journal.pone.0075407.g002

The estimation of linkage disequilibrium for each pair of SNPs presented strong linkage disequilibrium (D’ > 0.7) for two groups of markers (rs357552-rs357565-rs28485160 and rs2236405-rs28701981-2236407-rs10512248) (Table 3) [20]. Moreover, haplotypes were omitted from analysis if the estimated haplotype probabilities were less than 3% in either the HSCR or normal control subjects. We therefore adopted the haplotype distributions for these SNPs in the later analysis.

We chose only those haplotypes with significant frequency discrepancies between HSCR and control groups (Table 4) for presentation. As the frequencies were higher in the HSCR group compared to the control group, several haplotypes were found to be correlated with an increased odds ratio for HSCR: the haplotype A-G (rs2236405-rs2236407, P = 0.006, OR 2.21, 95% CI 1.25-3.90), the haplotype A-C-C-A-C-G-C (rs357552-rs357565-rs28485160-rs2236405-rs28701981-rs2236407-rs10512248, P = 0.004, OR 4.66, 95% CI 1.48-14.64), and the haplotype A-C-C-G-A-C-G-C (rs357552-rs357565-rs28485160-rs357564-rs2236405-rs28701981-rs2236407-rs10512248, P = 0.002, OR 5.52, 95% CI 1.66-18.35). Moreover, the haplotype analysis of these genetic markers revealed some significant global P values (Table 5). One seven-SNP-based haplotype was the most significant, giving a global P = 0.0036.

In the power calculations using the G*Power 3 program based on Cohen’s method [19], our sample size had > 80% power to detect a significant (α < 0.05) association for genotypes, alleles and haplotypes when an effect size index of 0.24 (corresponding to a “weak” gene effect) was used.

Discussion

The presence of a complete ENS is required for a normal functioning intestinal tract throughout its entire length. The mammalian ENS is derived from a small pool of progenitor cells known as the enteric neural crest cells (ENCCs) [21]. It is indicated that normal ENS formation depends on an appropriate balance between ENCCs proliferation, differentiation and migration during ENS development, which is also based on the coordination of different signaling pathways, and defects in any of the pathway components might lead to Hirschsprung disease [22-24].

The Hedgehog (Hh) signaling pathway is fundamental for numerous processes during embryonic development, and is also implicated in mediating proliferation and differentiation of ENCCs [13,25,26]. Extracellular Hh proteins bind to the receptor PTCH1 (a protein of 1447 amino-acid residues containing 12 transmembrane domains and two large extracellular loops) and restrain PTCH1-mediated inhibition of downstream signaling by the transmembrane protein smoothened (SMO) [27]. Releasing SMO activity further activates the downstream cascade by promoting the formation of the activator form of the GLI-Kruppel family member GLI1 family of zinc-finger transcription factor [13]. Thus, aberrant activation or function of PTCH1 is supposed to induce the improper activation of the Hh signaling pathway [12]. Ramalho-Santos and colleagues [28] suggested appropriate Hh function was necessary for the formation of sufficient enteric ganglia in the intestinal tract, whereas excessive activation of the Hh pathway in ENCCs can cause progenitor progression. In addition, mice lacking either the Sonic Hh- or Indian Hh-secreted proteins presented partial intestinal aganglionosis, accompanied by megacolon or ectopic ganglia formation [28].

In this study, we present the first indication that common genetic variants in the PTCH1 gene might confer altered susceptibility to Hirschsprung disease in the Han Chinese population and provide further support for the assumption that PTCH1 might be involved in the etiology of HSCR. The genetic analysis was performed using the Sequenom MassARRAY technology (Sequenom, San Diego, CA), and altogether 8 genetic polymorphisms, including two SNPs previously investigated by Ngan et al. [13] and six other SNPs, were genotyped in the 104 subjects with HSCR and the 151 normal controls. At two of the eight markers (rs357565 and rs2236405), there were statistically significant discrepancies of allele or genotype frequencies between HSCR and control groups. In particular, we found the C allele of rs357565 and the A allele and AA genotype of rs2236405 were more frequent in the HSCR subjects compared to normal controls (Table 2), implying that all might be the risk factors for Hirschsprung disease.

We further carried out haplotype analysis in the genetic markers with strong LD (D’ > 0.7) since haplotypes constructed from closely located SNPs will typically increase the statistical power for association with the disease. Our data presented some significant global associations with HSCR and the most significant window spanned seven markers, giving a global P = 0.0036 (Table 5). Moreover, the most significant haplotype A-C-C-G-A-C-G-C (rs357552-rs357565-rs28485160-rs357564-rs2236405-rs28701981-rs2236407-rs10512248, P = 0.002, OR 5.52, 95% CI 1.66-18.35) was five times as common in the HSCR group (6.4%) as in the control group (1.3%), indicating it might be a risk haplotype for HSCR, whereas the haplotype A-A (rs357552-rs357565) might be a protective factor against Hirschsprung disease since a lower frequency observed in the HSCR subjects (12.2%) compared to the controls (22.8%) (P = 0.003, OR 0.47, 95% CI 0.29-0.78).

Out of the two positive markers we found in the present study, rs2236405 is a missense variant located in the third exon of PTCH1 (Figure 1), and this SNP is an A>T transversion, changing a threonine residue to serine residue. The amino acid substitutions caused by missense variants are more likely to have serious consequences for the function or structural stability of the relevant protein [29,30]. Moreover, the other positive SNP, rs357565, is located in the untranslated region of PTCH1, and as non-coding regulatory variant rs357565 might also be functionally important, probably by affecting gene regulation and expression [29]. Ngan et al. [13] recruited canonical correlation analysis (CCA) to assess epistasis on data sets generated from a genome-wide association study on HSCR, and suggested that specific genotype constellations of PTCH1 and DLL3 (delta-like 3, a Notch ligand) SNPs conferred higher risk to HSCR. Rs357552 and rs10512248 included in the genotype constellations were also investigated in our study, and yet no positive results were observed at these two markers. However, since rs357565 is in strong LD with rs357552 (D’ = 0.96) (Table 3) and the correlated major alleles of both SNPs were implicated with a similar phenotype, we speculated that our findings might be regarded as an indirect support for the results observed by Ngan et al. [13].

In conclusion, our results provide a first indication that common genetic variants within PTCH1 might confer altered risk to Hirschsprung disease in the Han Chinese population, further supporting PTCH1 as a potential susceptibility gene for HSCR. If our findings can serve as a reference point for further replication studies in other ethnic groups and the mechanism can be verified in functional analyses, this could motivate genetic polymorphisms of PTCH1 as molecular markers to reach early diagnosis of clinical manifestations.

Acknowledgments

We appreciate the contribution of all the members participating in this study. We also would like to thank Dr. Guang He for her valuable suggestions and comments on the design of the study.