Three additional de novo CTCF mutations in Chinese patients help to define an emerging neurodevelopmental disorder

1 INTRODUCTION

CTCF gene (OMIM:604167) encodes the CCCTC-binding factor (CTCF). CTCF has been implicated in the formation of topologically associated domains (TADs) by binding to tens of thousands of DNA sequence-specific sites (CTCF motifs) in the human genome (Barutcu, Lian, Stein, Stein, & Imbalzano, 2017; Barutcu, Maass, Lewandowski, Weiner, & Rinn, 2018; Nora et al., 2017). It functions as a transcriptional insulator by preventing interactions between promoter and nearby enhancers and silencers (Austin, Lu, Majumder, Ahmed, & Boss, 2014; MacPherson & Sadowski, 2010; Nora et al., 2017; Xie et al., 2007); acts as a chromatin barrier by preventing the spread of heterochromatin structures and epigenetic modification (Chang et al., 2015; Guerra-Calderas et al., 2018; Hou, Zhao, Tanimoto, & Dean, 2008; Miura, Miyazato, Satou, Tanaka, & Bangham, 2018; Satou et al., 2016; Yang et al., 2003); and causes the bound DNA to form loops which influence the higher-order chromatin structure by forming CTCF homodimers (Battistelli, Busanello, & Maione, 2014; de Wit et al., 2015; Weth et al., 2010). CTCF is a multivalent transcription factor with 11 highly conserved zinc finger (ZF) domains, functions both as a transcriptional activator or a transcriptional repressor depending in the context of the specific binding sites and binding partners in the complex (Filippova et al., 1996; Holwerda & de Laat, 2013; Kim, Yu, & Kaang, 2015; Liu, Wu, & Wang, 2018; Rubio et al., 2008). Overall, CTCF plays important roles in regulating global chromatin organization and gene expression (Franco, Prickett, & Oakey, 2014). The involvement of the CTCF gene in human disease pathogenesis is emerging.

In 2013, Gregor et al. (2013) first identified a de novo frameshift variant by trio exome sequencing in CTCF in a patient with intellectual disability. Two more de novo variants were detected in the subsequent screening among a cohort of patients with intellectual disability. In 2017, Bastaki et al. (2017) reported a de novo frameshift variant in an Arab female patient with syndromic intellectual disability. So far, only four patients have been reported worldwide. Consistent clinical features among the four patients included developmental delay/intellectual disability, hypotonia, early feeding difficulty, microcephaly and mild facial dysmorphism. Based on the current evidence, the gene-disease association is classified as “Moderate” by the ClinGen gene curation protocol (https://search.clinicalgenome.org/kb/gene-validity/10143). Thus, this is an emerging disorder that requires additional genetic or functional evidence to establish the clinical validity between CTCF and the disease. The disease is defined as “Intellectual disability-feeding difficulties-developmental delay-microcephaly syndrome” by Orphnet (ORPHA:363611) and “Mental retardation, autosomal dominant 21” by OMIM (OMIM:615502).

Here we reported three unrelated Chinese patients with de novo CTCF variants and similar clinical phenotypes. Addition of these new cases can elevate the level of gene-disease association in ClinGen to “Strong” (see Section 4), thus firmly establishing the causal relationship between CTCF and the novel neurodevelopmental disorder. This report also expanded the mutation spectrum of CTCF gene and revealed the consistent phenotypes across populations and novel features of this condition.

2 MATERIALS AND METHODS

3 RESULTS

3.1 Case reports

All three Chinese patients were the first child of healthy parents without significant family history for genetic diseases. They all underwent uneventful full-term gestations. The main clinical features of the three Chinese patients have been listed in Table 1.

Table 1. Clinical features of Chinese patients with de novo CTCF mutations

	Patient 1	Patient 2	Patient 3
Variants in CTCF (NM_006565.3)	c.615_618delGAAA (p.Lys206Profs*15)	c.1699C > T (p.Arg567Trp)	c.329dupT (p.Gly111fs*29)
Gender	Female	Female	Male
Age at last examination	1 year and 7 months	8 years and 7 months	7 years and 11 months
Gestation	Full-term	Full-term	39 weeks
Birth weight	2.8 kg (<−1 SD)	2.65 kg (<−1 SD)	1.9 kg (<−3 SD)
Birth length	48 cm (<−1 SD)	50 cm (normal)	45 cm (<−3 SD)
Feeding difficulty	Yes	Yes	Uncertain
Muscular hypotonia	Yes	Yes, during childhood but improved afterward	No
Weight	8.25 kg (<−2 SD)	17.0 kg (<−2 SD)	18 kg (<−2 SD)
Height	76 cm (<−2 SD)	119.8 cm (<−3 SD)	120.3 cm (<−3 SD)
OFC	44 cm (<−2 SD)	48.2 cm (<−2 SD)	47.5 cm (<−3 SD)
Developmental delay/intellectual disability/movement delay	Yes; cognitive impairment	Yes; IQ 71; memory deficient; unable to jump with both feet	Yes; memory deficit; cognitive impairment
Age of walking	17 months	19 months	14–16 months
Age of first words	Language has yet to development	12 months	24 months
Behavior anomalies	Poor eye contact; irritability;	/	Sensitive; attention deficit; hyperactivity
Brain anomalies	/	MRI normal	CT normal
Recurrent infections	No	Yes	No
Abnormality of vision	Strabismus	Strabismus	Myopia
Hearing	Normal	Normal	Normal
Urogenital anomalies	Normal	Normal	Normal
Tooth development	No teeth development until 1 year and 7 months	Hypodontia; abnormal morphology of incisor; carious teeth	Hypodontia; caries
Facial dysmorphisms	Low set ears; cleft palate; upslanting palpebral fissures; unique eyebrows pattern; flat malar bone; micrognathia	Long eyelashes; unique eyebrows pattern; deep set eyes; upslanting palpebral fissures; strabismus; depressed nasal bridge; thin lips; flat malar bone; micrognathia	Widely spaced eye; depressed nasal bridge; upslanting palpebral fissures; flat malar bone; micrognathia
Ribs	11 pairs	12 pairs	12 pairs
Other anomalies	Palmar creases	Bone age delay for 2 years	Persistent patent ductus arteriosus, repaired at 8-year-old

Patient 1 was a 19-month-old girl who presented with developmental delay. At 22 weeks of gestation, an ultrasound examination showed that the fetus was 2 weeks smaller than gestation age. She did not present small for gestation age (SGA) at later gestation stages. She had hypotonia and feeding difficulty after birth. She developed postnatal growth retardation. She walked at 17 months and has yet to develop language. There was no dentition until 1 year and 7 months. She exhibited poor eye contact and strabismus. She also presented with irritability. Mild dysmorphic features included low set ears, high arched palate, as well as a unique eyebrow pattern (Figure 1a,d; see Section 4). In addition, the skeletal survey showed 11 pairs of ribs (Figure 2).

Patient 2 was an 8 years and 7 months girl who had a healthy younger brother. Parental heights were 170 cm for the father, and 160 cm for the mother. Her birth length (50 cm) and birth weight (2.65 kg) were within normal ranges. She exhibited hypotonia in early childhood but improved afterward. She started walking at 19 months. She still had difficulty jumping using both feet. She exhibited postnatal growth retardation (Figure 3) and delayed bone age (1 year and 4 months below chronological age). Growth hormone (GH) provocative test using arginine and l-dopa revealed a peak value of GH 3.58 ng/mL (1.95 ng/mL at 0 min, 3.58 ng/mL at 30 min, 2.79 ng/mL at 60 min, 0.71 ng/mL at 90 min, 1.88 ng/mL at 120 min, 1,96 ng/mL at 150 min, and 0.78 ng/mL at 180 min), suggesting a completely GH deficiency. Other hormonal measurements including luteinizing hormone (<0.1 IU/L), follicle stimulating hormone (0.53 IU/L), insulin-like growth factor 1(128 ng/mL), and insulin-like growth factor-binding protein 3 (2.68 mg/L) were also significantly lower than normal. From the age of 5 years and 7 months, she was started on recombinant human growth hormone (rhGH) treatment. She exhibited a good response to the treatment (Figure 3) during the first 2 years of the treatment. The growth velocity prior to treatment was 6.2 cm/year (from age 4 years 3 months to 5 years 7 months), her height at the start of treatment (age of 5 years 7 months) was 101.8 cm (−2.81 SD). The growth velocity of the first year treatment was 7.0 cm/year and the height increased 0.46 SD. The growth velocity of the second year treatment was 6.6 cm/year and the height increased 0.31 SD. By the end of the third year treatment (age 8 years 7 months), her height was 119.8 cm, the growth velocity dropped to 4.4 cm/year and the height decreased 0.12 SD (Figure 3). Her bone age was delayed for about 1 year at the end of the third year treatment. She underwent an intelligence test using the Chinese Wechsler Intelligence Scale for Children at the age of 8. Her Full-scale IQ was 71. She was particularly weak in numbering, math, and visual-spatial capability. Her verbal comprehension index was 81 and her processing speed index was 66. She had an amicable personality and good communication skills. She had an unusual interest in playing with crunchy and colorful papers. Her facial dysmorphic features included microcephaly, long eyelashes, and unique eyebrow patterns (Figure 1b,d), deep-set eyes, strabismus, depressed nasal bridge, thin lips and micrognathia. She had hypodontia and abnormal incisor. She had normal external genitalia.

Patient 3 was a 7 years and 11 months old boy at the time of the last examination. His mother had a history of spontaneous abortions. Oligohydramnios was noticed during pregnancy. His birth weight and birth length were below −3 SD. He was nursed in the incubator for 2 weeks after birth. It was not certain whether he had hypotonia and feeding difficulty after birth. He started walking at around 14 months of age and started talking at 2 years of age. He continued to exhibit growth retardation. He presented with attention deficit and hyperactivity, as well as irritability. He had poor memory and mild intellectual disability but had good social interactions. His facial dysmorphism included widely spaced eyes, depressed nasal bridge, severe tooth dysplasia, micrognathia, and an ear tag on the right side (Figure 1c,f).

3.2 Molecular analyses

Exome sequencing detected the following heterozygous variants in CTCF gene in probands respectively (RefSeq NM_006565.3): c.615_618delGAAA (p.Lys206Profs*15) in Patient 1, c.1699C>T (p.Arg567Trp) in Patient 2 and c.329dupT (p.Gly111fs*29) in Patient 3 (Figure 4). Sanger sequencing of parental samples confirmed that they were all de novo. Paternity was confirmed for all three cases. All three variants were absent from population databases (e.g., GnomAD). The variants detected in Patients 1 and 2 were previously reported in affected individuals, and the variant in Patient 3 is novel. They all can be classified as pathogenic following the ACMG/AMP guidelines (Richards et al., 2015). The variants appeared to distribute in multiple locations across the whole gene (Figure 4).

4 DISCUSSION

CTCF is an important regulator for the global genomic organization and gene expression (Franco et al., 2014). Somatic mutations of the CTCF gene had been detected in numerous cancers (Bailey et al., 2018; Chen et al., 2016; Klenova, Morse, Ohlsson, & Lobanenkov, 2002; Soto-Reyes et al., 2012; Umer et al., 2016). Germline mutations had been implicated in human disease (Zhou, Werelius, & Lindblom, 2004). Prior to this publication, two reports described a total of four individuals with syndromic intellectual disability and de novo variants in the CTCF gene (Bastaki et al., 2017; Gregor et al., 2013). Based on the available evidence, the gene–disease relationship regarding clinical validity is at a moderate level per ClinGen gene curation classification (https://search.clinicalgenome.org/kb/gene-validity/10143). In order to reach an affirmatory level, additional cases or functional evidence is needed. The new cases reported in this study provided sufficient evidence to re-classify the gene–disease relationship as “Strong” (the genetic evidence at the case level can reach a maximal score of 12). Given the fact that detected variants are distributed in the different regions of the gene and most of the variants are loss of function variants (Figure 3), we agree with Gregor et al. that the disease mechanism of this condition is haploinsufficiency (the haploinsufficiency score for CTCF gene can reach 3 based on ClinGen gene dosage sensitivity scoring protocol). Similar phenotypes were observed in the few individuals with deletion CNV involving the CTCF gene, which further supports this notion (Gregor et al., 2013; Hori et al., 2017) .

This condition has been named as “intellectual disability- feeding difficulties- developmental delay- microcephaly syndrome” by Orphanet (ORPHA:363611) based on the shared features of three patients described by Gregor et al. (2013). The consistent features now shared by the seven individuals are as follows: (a) neonatal hypotonia and feeding difficulties; (b) global developmental delay; (c) intellectual disability; (d) short stature; (e) microcephaly. In addition, we noticed that all our three patients have delayed or abnormal teeth development, presented with hypodontia and caries (Figure 5). This feature was also prominent in the Arab Patient [23]. The patient one reported by Gregor et al. (2013) also have abnormal teeth presentation, thus features related with teeth were observed in 5/7 patients. Strabismus and vision abnormalities were also recurrent (5/7). We noticed a unique pattern of eyebrow shared by most patients. As shown in Figure 1, the main characteristic of this unique eyebrow is that it spread out toward nasal root on the medial sides and the low edges of the eyebrow were positioned very close to the inner canthi. The eyebrow tapered off from middle-way to lateral two-third, so the overall morphology of the eyebrow is wider on the medial sides and thinner on the lateral sides. This eyebrow pattern is also visible from the patient photos published in the previous two papers. Long eyelashes were also a shared feature among our patients (Figure 1). Our patients also exhibited micrognathia, upslanting palpebral fissure, and flat malar bone (Figure 1). If these features can be validated in more patients, the condition can potentially be clinically recognized.

Yet the severity of the clinical presentation may differ significantly among patients, even between patients with the same mutation. Our Patient 1 carried the same mutation as the Arab patient (Bastaki et al., 2017). Our patient was born at full term without SGA whereas the Arab patient was delivered at 26-weeks of gestation with significant intra-uterine growth retardation and fetal stress. The Arab patient had a more severe motor delay, was unable to walk whereas our patient had normal motor function. The Arab patient needed the feeding tube whereas our patient had feeding difficulty but required no feeding tube. The congenital heart disease and external genitalia dysplasia observed in the Arab patient were absent from our patient. Our Patient 2 carried the same missense mutation as the Patient 3 reported by Gregor et al. (2013). Our patient had borderline intellectual disability whereas the latter had severe ID. Our patient has relatively preserved communication and language skills, while the latter had autistic features. Our patient showed more severe growth retardation.

CTCF affects genes and biological processes globally (Shen et al., 2015). The clinical consequence of CTCF gene mutations also seems to have a global effect on development and growth, both on neuronal (presented as microcephaly and intellectual disability; Hori et al., 2017; Lanni et al., 2013; Watson et al., 2014) and musculoskeletal (presented as short stature) tissues (Bailey et al., 2018; Lobanenkov & Zentner, 2018). Although variable, developmental delay and intellectual disability were mild among these Chinese patients. None of them exhibited autism presentations or significant behavioral issues, but they had attention problems, memory problems, and learning difficulties. The Patient 3 underwent rhGH treatment for 3 years, which showed a good response with no side effect. However, the effect did not last in the third year. With limited evidence, we do not have a good prediction regarding the prognosis of this condition, follow-up and more case studies are needed for better understanding this novel disorder.

In conclusion, we reported the clinical phenotypes of three unrelated Chinese patients with de novo heterozygous mutations in the CTCF gene. These new cases helped to establish the causal relationship and the clinical validity between the CTCF gene and this novel neurodevelopmental disorder. The presentations of developmental delay and growth retardation, as well as unique dysmorphism may help to define a clinically recognizable condition. rhGH treatment had some benefit but the long term effect is not known and the overall prognosis is uncertain.