A Novel MYCN Variant Associated with Intellectual Disability Regulates Neuronal Development

Dear Editor,

The V-MYC avian myelocytomatosis viral-related oncogene, a neuroblastoma-derived gene (MYCN, MIM: 164840) located on chromosome 2p24, was previously found to be associated with Feingold syndrome 1 (FGLDS1, MIM: 164280) [1]. FGLDS1 is an autosomal dominant disorder characterized by variable combinations of microcephaly, limb malformations, esophageal and duodenal atresias, and learning disabilities. Cardiac and renal malformations, vertebral anomalies, and deafness have also been described in a minority of patients [2]. Despite the involvement of intellectual disability in FGLDS1, the molecular mechanisms of the MYCN gene in regulating brain development remain largely unclear. Some truncated mutations in the N terminus of the MYCN have been identified in FGLDS1 [1, 3].

In this study, a Chinese girl affected by frontal bossing, hypertelorism, a high-arched palate, speech delay, and intellectual disabilities was recruited. We found a missense mutation in the MYCN gene, which explained the less severe symptoms of FGLDS1 but more symptoms of developmental disorder including intellectual disabilities. Thus, we propose to classify the patient as having non-canonical FGLDS1, and hypothesize that the missense mutation of MYCN contributes to the symptoms through alterations in brain development.

Whole-exome sequencing to a median of 50 × coverage identified 98671 genetic variants in the proband. After filtering, 448 rare variants in 433 genes were identified. A negative family history suggested de novo or autosomal recessive inheritance, and only variants located in morbid genes according to the Online Mendelian Inheritance in Man (OMIM) database and fitting in the inheritance model were identified as potential pathogenic mutations. Finally, a total of 3 variants in 3 genes were selected (Table 1).

Table 1.

Summary of morbid OMIM genes with de novo variants in the proband.

Gene	Chr	Location	NM	Mutation call	Hom/Het	Disease
MYCN	2	16082736	NM_005378	c.550G > T:p.Ala184Ser	Het	Feingold syndrome [MIM: 164280]
MEGF8	19	42857691	NM_001271938	c.3625C > T:p.Arg1209Trp	Het	Carpenter syndrome 2 [MIM: 614976]
PDE4D	5	58286737	NM_001165899	c.1006-8delCinsTC	Het	Acrodysostosis 2, with or without hormone resistance [MIM: 614613]

A de novo heterozygous variant in the MYCN gene (NM_005378, exon2, c.G550T, p.A184S) was identified in the proband. The MYCN mutation was further validated by Sanger sequencing. This MYCN A184S variant was not found in the parents, so it was defined as a de novo variant (Fig. 1A). Other two disorder-related genes were identified in the proband: MEGF8 [MIM: 604267] and PDE4D [MIM: 600129] (Table 1). The protein encoded by the MEGF8 gene [multiple EGF (epidermal growth factor)-like-domains 8] is a single-pass type I membrane protein that contains several EGF-like domains. Defects in this gene cause Carpenter syndrome, which is an autosomal recessive multiple congenital malformation disorder characterized by multisuture craniosynostosis and polysyndactyly, umbilical hernia, cryptorchidism, and congenital heart disease [4]. Genetic variations in PDE4D might be associated with susceptibility to stroke [5]. PDE4D is known to be associated with acrodysostosis 2 (ACRDYS2) [MIM: 614613], a pleiotropic disorder characterized by skeletal, endocrine, and neurological abnormalities. The skeletal features include brachycephaly, midface hypoplasia with a small upturned nose, brachydactyly, and lumbar spinal stenosis. The endocrine abnormalities include hypothyroidism and hypogonadism in males and irregular menses in females [6, 7]. The main clinical manifestations of the MEGF8 and PDE4D genes were not consistent with those of our proband, so we focused on the variant in the MYCN gene as the most likely disease-causing event.

Fig. 1.

Expression pattern of MYCN in mouse brain and its role in neural development. A Known nonsense mutations of the MYCN gene that cause FGLDS1 (green and black) and the newly-found mutation (red). Sanger sequencing results confirmed that the mutation was de novo. B Western blots of MYCN protein levels in the cortical region of P0-P35 mice. C, D Quantitation of PTEN (C) and P53 (D) levels in HEK293 cells transfected with control (CTRL), MYCN, and MYCN-A184S plasmids. RNA samples were collected 48 h after plasmid transfection, and levels were assessed by qPCR. *P < 0.05, **P < 0.01, ***P < 0.001. E Quantitation of PTEN levels in mouse primary cortical neurons transfected with CTRL, MYCN, and MYCN-A184S plasmids. RNA samples were collected 5 days after plasmid transfection. **P < 0.01. F, G Representative images (F) and statistics for axonal length (G) of mouse primary cortical neurons transfected with CTRL, MYCN-WT, and MYCN-A184S. Neurons were fixed and immunostained with GFP antibody at 3 DIV for measurement of axon length. More than 70 examples in each group of neurons were randomly selected for counting. ***P < 0.001. H Quantitation of MYCN mRNA levels in mouse cortical neurons transfected with sh-DsRed-GFP (CTRL) and MYCN-shRNA (sh-MYCN); mRNA levels were assessed by qPCR. RNA samples were collected 5 days after plasmid transfection. **P < 0.01. I Representative images of mouse primary cortical neurons transfected with sh-DsRed-GFP (CTRL) or sh-MYCN. Neurons were fixed and immunostained with GFP antibody at 3 DIV. J Measurements of axonal length in the sh-DsRed-GFP (CTRL) and sh-MYCN groups. More than 50 examples from each group of neurons were randomly selected for counting. ***P < 0.001.

To investigate the role of the MYCN protein during neuronal development, we first examined the expression pattern of MYCN during mouse brain development. We found that the MYCN protein was consistently expressed during postnatal days 0–35 in the cortical region (Fig. 1B). MYCN is a critical tumor-related gene and may be associated with numerous tumor suppressors, including p53 and PTEN [8]. Thus, we cloned the human wild-type MYCN gene and added the A184S mutation to investigate the function of MYCN in neuronal development.

First, we determined whether MYCN regulated p53 and PTEN expression by overexpressing MYCN in human embryonic kidney (HEK293) cells and mouse primary cortical neurons. Interestingly, we found that MYCN significantly repressed the expression of p53 and PTEN in the HEK293 cells at the mRNA level (Fig. 1C, D). Furthermore, we verified at the protein level that overexpression of MYCN suppressed PTEN as well (Fig. S1A, B). Importantly, the A184S mutation of MYCN significantly inhibited PTEN expression in mouse cortical neurons, while the wild-type (WT) MYCN did not, suggesting that A184S is a gain-of-function mutation for MYCN (Fig. 1E).

Next, we investigated the roles of WT and A184S-mutated MYCN in neuronal development. Surprisingly, we found that MYCN overexpression and the A184S MYCN mutation significantly enhanced the length of axons (Fig. 1F, G), further suggesting that MYCN plays a critical role in regulating the development of neuronal morphology and that A184S is likely to be a gain-of-function mutation.

Then we designed short-hairpin RNA against the mouse Mycn gene to confirm the role of Mycn in neuronal development (Fig. 1H). The shRNA against Mycn was transfected along with the vector into mouse primary cortical neurons and the neuronal morphology was observed after 2–3 days. Consistently, knockdown of Mycn strongly inhibited axonal and neurite growth (Fig. 1I, J), indicating that Mycn plays a bi-directional role in regulating neuronal morphology.

To further define the functional differences between the WT and the A184S MYCN mutation, we collected HEK293 cell samples transfected with MYCN and MYCN A184S vectors and sent them for RNA sequencing. Significant differences in the expression patterns between the WT and the A184S group were found (Fig. 2A). Interestingly, Gene Set Enrichment Analysis (GSEA) showed that the A184S mutation group had greater activation in several pathways, such as the axon guidance, PI3K-AKT activation, and Wnt signaling pathways (Fig. 2B). We then analyzed the differences in microRNA expression between the two groups (Table S1) and obtained predicted target genes using the Miranda and RNAhybrid analysis methods. The target genes were further identified using the Fisher analysis method to classify pathways using Gene Ontology analysis. Notably, the axon guidance, Wnt signaling, and PI3K-AKT signaling pathways were among the top 15 (Fig. 2C, D), suggesting that the A184S mutation has an enhancing effect on axonal growth.

Fig. 2.

mRNAs and microRNAs were differentially expressed in the MYCN-A184S group compared to HEK293 MYCN-WT cells. A Heatmap of differentially-expressed mRNAs between the MYCN-A184S and MYCN-WT groups; each group contained four samples (false discovery rate < 0.05, fold change > 2). B Gene set enrichment analysis of the axon guidance, PI3K-AKT activation, and Wnt signaling pathways in the MYCN-A184S and MYCN-WT groups at the mRNA level. C Pathway analysis of target genes predicted by differentially-expressed miRNAs in the MYCN-A184S group compared to the MYCN-WT group. D Gene ontology (GO) analysis of target genes predicted by differentially-expressed miRNAs in the MYCN-A184S group compared to the MYCN-WT group.

In this report, we present a female patient with frontal bossing, hypertelorism, a high-arched palate, speech delay, and mental retardation. The patient carries a transition of c.G550T, which causes a heterozygous mutation of p.A184S in exon 2 of the MYCN gene. Furthermore, this is a de novo mutation validated by Sanger sequencing. Until now, all of the identified mutations mapped within the repressor domain have been nonsense mutations (E73X, W77X, G151X, and S221X) or frame-shift mutations (E47fs, G101fs, and P228fs), which would remove most regions of the wild-type protein, as in haploinsufficiency, which induces FGLD1 [1, 3, 9, 10]. However, we identified a missense mutation of p.A184S in this region. This mutation is a milder functional defect than haploinsufficiency, so it may contribute to a less severe and atypical phenotype of FGLD1.

In total, 488 rare variants in 433 genes were identified in the proband after the filtering step. We focused on the variants that were located in the morbid OMIM genes and fitted into the inheritance model in this three-person family without any known family history. Although genes without clear associations with the specific phenotype may contribute to the clinical manifestations of the proband, such variants need more comprehensive evidence to confirm their relevance.

MYCN is a downstream target of Shh signaling, which promotes the rapid cell division of granule neuron progenitors (GNPs) in mice [11]. Overexpression of MYCN promotes the proliferation of GNPs independent of Shh signaling, and conversely, its conditional deletion during embryonic cerebellar development results in severe GNP deficiency, perturbation of foliation, and reduced cerebellar mass [11]. Valli et al. have shown that a functional axis between MYCN and cyclin-dependent kinase-like 5 governs both the neuron proliferation rate and differentiation, potentially contributing to the neurological symptoms in patients with MYCN mutations [12].

Maintaining the normal morphology of neurons is crucial for basic brain functions, since axons play important roles in structural formations and signal transmission. Dysfunctional axons may cause many neurological disorders, such as autism spectrum disorders [13]. In this study, we focused on the role of dysfunctional axons in the intellectual disability found in FGLDS1, and found that overexpression of MYCN strongly promoted axonal growth, whereas its knockdown repressed the axonal growth of mouse cortical neurons. Interestingly, the A184S mutation in MYCN identified in this patient with non-canonical FGLDS1 syndrome had a stronger effect on axonal growth.

Furthermore, we found that the A184S mutation significantly inhibited the expression of p53 and PTEN, according to the RNA sequencing results. The GSEA pathway and GO analysis indicated that the A184S mutation has a greater effect on the axonal guidance, PI3K-AKT activation, and Wnt signaling pathways. PTEN is a known important component in neurite growth via regulating the PI3K-AKT-mTORC1 signaling pathway, and its downregulation promotes neurite growth [14], which was also shown in our results. The Wnt receptor in Wnt signaling is also associated with axonal guidance [15]. These findings suggest that A184S is a gain-of-function mutation.

Although further research is still needed, our study suggests a possible role of the A184S mutation in intellectual disability. The results shed new light on the mechanistic insights of MYCN-related brain developmental disorders and provide a potential candidate for mutation screening in related brain disorders.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no competing financial interests.