Main

Sotos syndrome (SoS, OMIM 117550) is an overgrowth disorder characterized by distinctive facial features, craniomegaly, advanced bone age, early and excessive growth with accompanying neurodevelopmental and cognitive delay, and intellectual disability (1, 2). Characteristic facial features include a prominent forehead; sparse frontal hairline; a sharp, pointed mandible; and downward-slanting eyes. Learning disability can range from mild to severe, with difficulty in speech and communication. Growth is usually premature, with the birth length, head circumference, and bone age ranging above the 90th percentile (3, 4).

This syndrome is usually accompanied by other clinical problems such as spine deformities, and cardiac and genitourinary disorders (3). Cardiac anomalies occur in ~20% of patients, and these may include patent ductus arteriosus, atrial septal defect, or ventricular septal defect. Ventricular dilatation and vesicoureteral reflux are the most common cranial and renal abnormalities, respectively (4). Other symptoms or features associated with SoS range from constipation to hearing or vision loss, cryptorchidism, hernias, hydrocele, gastroesophageal reflux, and many others (4).

Mutations of NSD1 are the known etiologic factors for this disease entity, which may include frameshift and nonsense mutations, partial gene deletions, and microdeletions. NSD1 molecular genetic screening is used to confirm any suspected case of SoS (4). However, additional confounders, such as APC2 and the MAPK/ERK signaling pathway, have recently been identified as risk factors for SoS development based on genome-wide levels discovered through whole-exome-sequencing (WES) and genome-wide expression studies (5, 6).

To the best of our knowledge, Hirschsprung’s disease (HSCR) has not been documented as part of SoS. Whether the combined phenotype of SoS with HSCR is sporadic or associated with the syndrome is yet to be discovered. Our objective is to report on a case of SoS associated with HSCR, perform a genetic analysis, and compare the results with genetic variations in our previous genome-wide association study (GWAS) of HSCR (7).

Results

Case

A 2-year-old boy, who was known to have SoS, visited our Seoul National University Bundang Hospital as he was suffering from constipation since birth. Bowel movements were painful and frequency varied from once a day to once in 2 days. According to the birth record, he was born at 37±2 weeks age of gestation via normal spontaneous vaginal delivery weighing4.07 kg, and he had a history of polyhydramnios and neonatal jaundice. There was passage of meconium within 24 h of birth, with irregular bowel movements.

In addition, the boy had a history of feeding problems (such as abdominal distension aggravated by feeding, frequent small volume of loose stool-like incontinence, and occasional projectile vomiting) and developmental milestones were delayed. He also experienced poor head control since infancy. Speech and walking were observed at 2 years of age. Past medical history included cryptorchidism on the right side, for which he underwent orchidopexy. An echocardiogram was not completed because of patient restlessness and noncooperation. However, brain magnetic resonance imaging showed prominent extra-axial fluid collection in the frontal convexities bilaterally with focal hemosiderin deposit in the left cerebellar hemisphere.

Physical examination showed typical facial features, with prominent forehead; macrocephaly; receding hairline; down-slanting palpebral fissure; long, narrow face; and long chin. The height was 0.98 m (>95th percentile) and weight was 17.0 kg (>95th percentile). The abdomen was soft and markedly distended, with palpable soft to firm masses and abundant fecal material on colonic irrigation.

The patient underwent a laparoscopic-assisted trans-anal pull-through Soave procedure with rectal biopsy on frozen sections. There was an adequate number of normal ganglion cells on histopathologic frozen biopsy of the upper rectum. The patient started a liquid diet on the first postoperative day. The diet was progressed, and the rest of the patient’s hospital stay was unremarkable. Final histopathologic examination of the resected bowel showed an aganglionic segment 13 mm distally and a ganglionic segment 87 mm proximally (Figure 1), with no identified immature ganglion cells, submucosal and myenteric plexus hyperplasia, or giant ganglia.

Figure 1
figure 1

Histological features of HSCR diagnosis. (a) The presence of ganglion cells on hematoxylin and eosin (H&E) staining. Ganglion cells of proximal segment are indicated by arrow. (b) No ganglion cell on H&E staining of the distal segment. (c) CD56 positive and (d) MAP2 negative on immunohistochemistry staining of the distal segment.

Full size image

Genetic Analysis

In the previous genetic test for SoS diagnosis of the patient, two mutations of NDS1, c.2465C>A (p.Ser822Tyr) and c.4347T>A (p.Cys1449*), were observed. The c.4347T>A (p.Cys1449*) mutation, leading a novel nonsense mutation in NSD1, was assumed to be a cause for this clinical syndrome, whereas both parents of the patient tested negative for the mutation (Figure 2a). In the case of c.2465C>A (p.Ser8225Tyr), this variant was also identified as novel, as it was not found in the dbSNP database of NCBI (http://www.ncbi.nlm.nih.gov/snp/), and there was a negative genotype in the father and a positive genotype in the mother (Figure 2b).

Figure 2
figure 2

Resequencing in the patient and his parents. (a) NSD1 c.4347T>A (p.Cys1449*) and (b) c.2465C>A (p.Ser8225Tyr). Primer pairs (5′-ccctgtctgttggagcattt-3′ and 5′-CAGATTTTGGGCAAACCAGT-3′ for c.4347T>A; 5′-TGTTCACAAACCCCAGTCAG-3′ and 5′-TTGATGGCTTTGATGTTCCA-3’ for c.2465C>A) are used. WT, wild type.

Full size image

In this study, WES was performed at a mean coverage of 87 × for the patient with SoS associated with HSCR. Among a total of 12,219 missense and nonsense variants identified, filtering (see detailed description in the Methods section) ultimately revealed 58 variants in 56 genes (Table 1). In order to find a relationship between SoS and HSCR, further comparison between WES in this study and our recently meta-analyzed GWAS of HSCR carried out across Chinese, Korean, and European populations was performed. Our results showed that mutations in several well-known HSCR-related genes (RET, NRG1, NRG3, EDNRB, GDNF, NRTN, SOX10, and SEMA3) and variations with additional strong signals observed in our previous GWAS (SLC6A20, IL-11, SLC6A20, and VAMP5) (7, 8, 9, 10) were not found in the patient with SoS associated with HSCR. However, among the genes showing variants in WES, three (ZNF827, FGD2, and KCNJ12) that had shown significance for HSCR in our previous GWAS were overlapped.

Table 1 Mutations detected by whole-exome sequencing
Full size table

Discussion

This report is the new occurrence of HSCR that has not previously been reported in SoS. As shown in this case, the occurrence of HSCR could be purely co-incidental and may not be a feature of SoS. The incidence of HSCR (1.4–2.8 in 5,000 live births) varies across ethnic groups, with the highest incidence in Asians (11). In the patients with SoS, gastrointestinal symptoms, including constipation (>15% of cases) as a common complaint, have also been reported (1, 12, 13). This clinical feature could hypothesize that sometimes the constipation observed in SoS could have a component of poor ganglion formation.

The diagnosis of SoS is made on the basis of history and physical examination, which includes the cardinal features of typical facial dysmorphism; increased velocity of growth and weight with a relative development rate of >2 or the 75th percentile; advanced bone age; and some form of developmental delay such as in talking, walking, or other developmental milestones (1, 2, 3, 14). These cardinal features are present in >90% of patients with SoS. Other symptoms associated with SoS encompass cognitive problems, mild-to-severe intellectual disability, behavioral problems, and cardiac problems. These problems may include enlarged ventricles and other congenital heart diseases; neurologic abnormalities such as seizures; increased subarachnoid spaces, or even pituitary and hypothalamic tumors; and genitourinary problems such as vesicoureteric reflux disease, cryptorchidism, or dysplastic kidneys (1, 2, 3, 5). Although all these disease entities have been associated with SoS, anorectal malformations have not been documented. Research and retrospective data have found that HSCR with associated anomalies or syndromes has poor outcomes such as long hospital stays and long time for recovery (15).

Genetic analyses of SoS have revealed mutations of NSD1 in 80% of affected patients, and it is believed to be an autosomal dominant disorder (14, 16). Various mutations, including nonsense, frameshift, and microdeletion in chromosome 5q35, have been detected in SoS patients (12, 14, 16, 17, 18). In the case of HSCR, many confounding factors, including the major RET mutations, are known to be associated with HSCR susceptibility. In addition, HSCR has been shown to be associated with other syndromes involving genetic mutations or chromosomal anomalies (11). Among the identified variations in this study, p.R261H and p.I262S of KCNJ12 are common variants with a population frequency of 0.163 and 0.178, respectively, suggesting that they may have little possibility to affect this rare disease. In the case of other overlapping genes (ZNF827 and FGD2), no literature clues related to SoS could be found. Although the functions of these genes are little understood, they have been reported to have roles in hepatic traits and leukocyte signaling (19, 20). On the other hand, considering that constipation is commonly seen in SoS, future studies for the possibility that the overlapping genes could relate to constipation or HSCR are required.

In additional gene ontology (GO) analysis of the detected genes with variants in the patient using the WEB-based GEne SeT AnaLysis Toolkit (http://bioinfo.vanderbilt.edu/webgestalt/), two genes (BRCA1 and TUBGCP6) involved in gamma-tubulin-related categories (GO:0008274, GO:0000931 and GO:0000930) were predicted to have a significant value in SoS associated with HSCR (minimum P=0.0002; adjP=0.01 for multiple test adjustment). Considering that tubulins function as a component of the enteric nervous system and show differential expressions along the entire colonic tract related to HSCR (21, 22), further studies of gamma-tubulin may be valuable in understanding SoS-associated gastroenterological diseases.

Recognizing HSCR early, especially in an SoS patient, is a significant part of management of the syndrome. Moreover, constipation in SoS patients may point to the possibility of an associated HSCR. To our knowledge, this is the first case report regarding the association of the two disease entities. Treatment of these individual diseases requires a multidisciplinary approach and collaborative treatment plan. The recognition of the genetic etiology may hopefully lead to the possibility of screening, as well as raising awareness that SoS and HSCR may coexist. In conclusion, HSCR should be considered as part of the clinical and diagnostic workup in SoS patients with constipation, as untreated or undiagnosed HSCR may contribute to severe morbidity. Further studies are also required to determine whether there is a genetic relationship between SoS and HSCR.

Methods

This study was approved by the Institutional Review Board (IRB number B-1504-294-702) of Seoul National University Bundang Hospital. After the patient’s guardian provided written informed consent, the patient’s medical records were reviewed.

In this study, we performed a genetic analysis using WES to investigate genetic variations and compared them with our previous GWAS of HSCR. Isolated, genomic, blood DNA from the SoS patient was used for exome library construction using the Ion AmpliSeqTM Library Kit 2.0 (Life Technologies, Waltham, MA). Library templates were then sequenced using the Ion PI Chip (Life Technologies) base on the Ion Proton Sequencer System (Life Technologies) according to the manufacturer’s instructions.

After processing sequencing reads using the Ion Torrent Suite software v 4.0.2 (Life Technologies), high-quality sequencing reads were mapped to the complete hg19 human genome (UCSC version, February 2009). On the basis of the variant discovery using Torrent Variant Caller v 4.2 (Life Technologies), significant variants were identified as follows: (i) mutations predicted to be positioned at frameshift, nonsense, or essential splice sites or (ii) missense variants predicted to affect the function with SIFT score (http://sift.bii.a-star.edu.sg) <0.05 and Polyphen2 score (http://genetics.bwh.harvard.edu/pph2) >0.85. Then, known common variants with a frequency over 0.01 in 1000 Genomes Project data were excluded.