Association between DSCAM polymorphisms and non-syndromic Hirschsprung disease in Chinese population

Hirschsprung disease (HSCR) is the most common, complex genetic disorder of the enteric nervous system (ENS) with an incidence of one in 5000 live births [1, 2]. It is characterized by the absence of enteric ganglia in the myenteric and submucosal plexuses along a variable length of the hindgut [3]. HSCR can occur in combination with other conditions, such as Waardenburg syndrome, Mowat-Wilson syndrome; or Downs syndrome which are described as syndromic. HSCR occur without other conditions were referred to as non-syndromic [2, 4]. Clinically, according to the length of the aganglionic segment, HSCR can be classified as short-segment HSCR (S-HSCR, 80% of cases) and long-segment HSCR (L-HSCR, 20% of cases) [5]. In few L-HSCR cases, the entire colon and even the whole bowel are involved and can be further classified into Total Colonic aganglionosis (TCA) and total intestine aganglionosis (TIA) (together less than 5% of cases). L-HSCR patients with longer aganglionic segments have been recognized as presenting poorer life quality and when compared to S-HSCR patients [6]. However, the etiology of the disease and its subclinical manifestations was still uncertain.

More than 15 loci have been identified as contributing to Hirschsprung disease, [7]. Most of them were involved in Enteric Neural Crest Cell (ENCC) proliferation, maintenance of Enteric nervous system (ENS) progenitors and neuronal or glial differentiation and hence only indirectly influence migration [7]. They were roughly classified into four major groups: RET and its ligands; EDNRB and EDN3 and ECE1 in the same pathway; the NRG homologs (NRG1 and NRG3);and the SEMA signaling pathway (SEMA3C and SEMA3D) [2, 4]. Syndromes associated with HSCR are usually considered as monogenic diseases; Waardenburg syndrome type 4 (Waardenburg-Shah syndrome) was considered as 100% coincident with HSCR. Down syndrome (trisomy 21) is the most common chromosome abnormality associated with HSCR which occurs in 2-10% of all individuals with HSCR [8]. In contrast, approximately 3% of individuals with Down syndrome have HSCR [9].

Jannot et al. [10] revealed two common variants (rs2837770 and rs8134673) in DSCAM as associated with 26 Caucasian HSCR patients accompanied with Down syndrome. DSCAM is expressed at a high level in the developing nervous systems in human, mouse and drosophila, implying a crucial role for DSCAM in the development of the nervous system [11]. However, they failed to replicate the association using 220 non-syndromic HSCR trios and public data awaiting for further replication.

Hence, in order to identify the association of DSCAM with non-syndromic HSCR, we conducted a replication study on the two SNPs in DSCAM using 1394 sporadic HSCR cases and 973 controls in South Chinese population. Further experiment was still required to further digest its functional mechanism to both central nervous system (CNS) and ENS.

The exact cause of HSCR is unknown. In an attempt to confirm the association of DSCAM with HSCR, we collected 1394 patients and 943 controls in South Chinese population, and successfully replicated two SNPs in DSCAM as associated with non-syndromic HSCR, especially in patients affected with a short anglia segment. We also observed a distinct lower expression in aganglionic segment compared with ganglionic segment and normal controls. Of course, the underlying mechanism to HSCR is still unclear which requires for further examination.

DSCAM has been repeatedly reported as a crucial player in neural development in vertebrates. It mediates homophilic attraction in neuronal layer-targeting [15]. Recently DSCAM has been identified as another receptor for netrin-1 [16]. It also works as a receptor for netrin in mediating axon guidance [17]. In addition to the contribution to DS-HSCR patients demonstrated by Jannot et al. [10], our results point to DSCAM as a susceptibility locus to non-syndromic HSCR patients. We also observed that the association of the two replicated SNPs mainly come from the effect on S-HSCR patients, which is similar to the associated common variants in RET mainly affecting the sporadic S-HSCR patients [18]. They also found rare mutations in RET were more likely to affect the severe cases including L-HSCR and TCA cases. At this stage, the predisposition of DSCAM to S-HSCR may due to the limited samples for L-HSCR patients and TCA patients, we don’t have enough power to detect the hidden association signals. It is also possible the common variants only play roles with S-HSCR patients with moderate effect, and there may exist rare variants which serve as casual mutations to the severe cases of the disease. Further study is required to validate the association in an independent cohort or directly search the potential functional rare variants by target region sequencing on DSCAM.

The lower expression of DSCAM was observed in the aganglionic segment of HSCR in current study. However, there still exist a gap how the associated SNPs may regulate the expression of DSCAM. We failed to identify the potential roles for the two SNPs through RegulomDB annotations [19]. Further experiments were required to further digest how DSCAM may affect the disease status of HSCR.