Article Text
PDF
Abstract
Background In this study, we aimed to identify the gene abnormality responsible for pathogenicity in an individual with an undiagnosed neurodevelopmental disorder with megalencephaly, ventriculomegaly, hypoplastic corpus callosum, intellectual disability, polydactyly and neuroblastoma. We then explored the underlying molecular mechanism.
Methods Trio-based, whole-exome sequencing was performed to identify disease-causing gene mutation. Biochemical and cell biological analyses were carried out to elucidate the pathophysiological significance of the identified gene mutation.
Results We identified a heterozygous missense mutation (c.173C>T; p.Thr58Met) in the MYCN gene, at the Thr58 phosphorylation site essential for ubiquitination and subsequent MYCN degradation. The mutant MYCN (MYCN-T58M) was non-phosphorylatable at Thr58 and subsequently accumulated in cells and appeared to induce CCND1 and CCND2 expression in neuronal progenitor and stem cells in vitro. Overexpression of Mycn mimicking the p.Thr58Met mutation also promoted neuronal cell proliferation, and affected neuronal cell migration during corticogenesis in mouse embryos.
Conclusions We identified a de novo c.173C>T mutation in MYCN which leads to stabilisation and accumulation of the MYCN protein, leading to prolonged CCND1 and CCND2 expression. This may promote neurogenesis in the developing cerebral cortex, leading to megalencephaly. While loss-of-function mutations in MYCN are known to cause Feingold syndrome, this is the first report of a germline gain-of-function mutation in MYCN identified in a patient with a novel megalencephaly syndrome similar to, but distinct from, CCND2-related megalencephaly-polymicrogyria-polydactyly-hydrocephalus syndrome. The data obtained here provide new insight into the critical role of MYCN in brain development, as well as the consequences of MYCN defects.
- missense mutation
- neurogenesis
- neurodevelopment
- neuroblastoma
- polydactyly
Statistics from Altmetric.com
Introduction
The MYC transcription factor family targets proliferative and apoptotic pathways, and members of this family, MYC, MYCL1 and MYCN, are well-established proto-oncogenes in humans.1 MYCN amplification is found in about 25% of neuroblastoma cases, the most common malignant extracranial solid tumour in childhood.2 MYCN also plays an important role in the early embryonic development of various organs, including the central nervous system, limb bud, lung, gut and heart.3–6
MYCN mRNA expression was reported to be high in fetal, but less so in adult, human brain.7 Germline heterozygous loss-of-function mutations or microdeletions in MYCN are known to cause Feingold syndrome (MIM: 164280), which is characterised by microcephaly, learning disabilities and limb malformations.8 9 In the cerebellum, Mycn (murine orthologue) activity regulates granule neuron proliferation through induction of Ccnd1 and Ccnd2 (murine orthologues of human CCND1 and CCND2, respectively).10 While Mycn overexpression was shown to promote proliferation of granule neuron progenitor cells in the cerebellum, its conditional loss of Mycn resulted in impaired proliferation of granule neuron progenitor cells, leading to reduced cerebellar mass.4 Furthermore, CCND2 gene abnormalities that impart excessive protein stability cause megalencephaly-polymicrogyria-polydactyly-hydrocephalus syndrome (MPPH; MIM: 615938).11 Notably, CCND2 stability also appears to be augmented by other MPPH-associated mutations in AKT3, PIK3R2 and PIK3CA.12
Phosphorylation of MYCN is crucial for its structural stability. It is stabilised and activated through phosphorylation of Ser62 (S62), which is followed by priming phosphorylation at Thr58 (T58).13 14 Dephosphorylation of S62 then sensitises T58-phospho-MYCN to interact with ubiquitin ligases such as F-box and WD repeat domain-containing 7 (FBW7), leading to proteasomal degradation.13 15 16 S62 and T58 residues are highly conserved across all MYC family members, and frequently mutated in c-Myc in Burkitt’s lymphoma, consistent with their functions in cell proliferation and differentiation.17
Here we report a de novo heterozygous missense mutation in MYCN identified by whole-exome sequencing of a Japanese boy with an intellectual disability (ID), distinctive facies, megalencephaly, ventriculomegaly, hypoplastic corpus callosum, postnatal growth retardation, postaxial polydactyly and neuroblastoma. Biochemical and cell biology experiments revealed that the mutation renders MYCN resistant to proteolysis and may improperly potentiate cortical neuron proliferation. We conclude that this mutation functions in a gain-of-function manner rather than the previously reported loss-of-function mutations that cause Feingold syndrome.
Methods
Whole-exome analysis
Molecular diagnosis was performed using whole-exome sequencing on a patient with an increased head circumference and neurological symptoms, but with no mutations in 15 known megalencephaly genes.18 The genomic DNA was extracted from peripheral blood, partitioned using the SureSelect XT Human All Exon V6 capture library (Agilent Technologies, Santa Clara, California), and DNA sequencing was performed using 150 bp paired-end reads with an Illumina HiSeq 4000 sequencer. To identify disease-causing mutations, we excluded variants with a minor allele frequency >0.5% in public databases and an inhouse control database (1044 controls of normal Japanese individuals and with other diseases), except those also identified as pathogenic mutations in the National Center for Biotechnology Information (NCBI) ClinVar and Human Gene Mutation Database (HGMD) databases. The public databases used were dbSNP150; 1000 Genomes Project; Exome Aggregation Consortium (ExAC); National Heart, Lung, and Blood Institute Exome Sequencing Project 6500; Human Genetic Variation Database; and the Integrative Japanese Genome Variation. We then excluded non-functional mutations to select non-synonymous single nucleotide variants (SNVs), insertions and deletions, and splice site variants. The mutation was confirmed by Sanger sequencing of PCR-amplified products.
Plasmids
Mouse Mycn was amplified by RT-PCR from pooled adult mouse brain RNA, and the cDNA was cloned into the pCAG-Myc vector (Addgene, Cambridge, Massachusetts). pCDNA3-hemagglutinin (HA)-tagged human MYCN was obtained from Addgene (plasmid #71463). Site-directed mutagenesis (KOD-Plus Mutagenesis Kit, Toyobo, Osaka, Japan) was then used to generate Mycn-T58M, MYCN-T58M and MYCN-P44L mutant constructs from pCAG-Myc-Mycn and pCDNA3-HA-MYCN, respectively. Mouse Fbw7 cDNA, a kind gift from Professor K Nakayama (Kyushu University, Fukuoka, Japan), was cloned into the pCAG-Myc vector. All constructs were verified by DNA sequencing.
Cell culture
HEK293T cells were cultured and transfected with Lipofectamine 3000 (Life Technologies Japan, Tokyo) according to the manufacturer’s guidelines. Mouse neuronal stem cells were isolated from embryonic day (E) 14 embryos and kept in the growth medium with basic fibroblast growth factor, B27 and epidermal growth factor (Thermo Fisher Scientific) supplement. Primary neurospheres were dispersed with 0.2% trypsin, and cells were electroporated with a CUY21 electroporator (Nepa Gene, Chiba, Japan), according to the manufacturer’s guidelines.
Preparation of mouse whole-brain extracts
Brains at various developmental stages were homogenised with 10 volumes of 50 mM Tris-HCl buffer (pH 7.5) containing 0.1 M NaF, 5 mM EDTA, 2% sodium dodecyl sulfate (SDS), 10 µg/mL aprotinin and 10 µg/mL leupeptin.19 Protein concentration was estimated with a micro bicinchoninic acid protein assay kit (Thermo Fisher Scientific) with bovine serum albumin as a standard.
Western blot
Indicated amounts of tissue or cell extracts were separated by sodium dodecyl sulfate-poly acrylaide gel (SDS-PAGE) (10% gel) and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, Massachusetts). After blocking with 5% skim milk powder, membranes were incubated with primary antibodies against N-Myc (sc-53993; Santa Cruz Biotechnology, Texas, USA), phospho-c-Myc (pThr58) (ab28842; Abcam, Cambridge, UK), phospho-n-Myc (S54) (Bethyl Labs, Montgomery, Texas), CCND1 (sc-450; Santa Cruz Biotechnology), CCND2 (ab 3085; Abcam), GFP (sc-9996; Santa Cruz Biotechnology), Myc-tag (#562; MBL, Nagoya, Japan) and GAPDH (#5174; Cell Signaling Technology, Danvers, USA), followed by incubation with horseradish peroxidase-conjugated secondary antibody (GE Healthcare, Little Chalfont, UK). Densitometric quantification was performed using the ImageJ software.
In utero electroporation and EdU incorporation experiments
In utero electroporation was performed with pregnant Institute of Cancer Research (ICR) mice as previously described,20 with the following amendments. Embryos were electroporated with pCAG-EGFP together with pCAG-Myc-Mycn or the pCAG-Myc control vector at E14.5. 5-ethynyl-2’-deoxyuridine (EdU) injection (25 mg/kg body weight) was performed at 24 hours, then at 48 hours brains were processed for sectioning and staining for EdU, Ki67 and green fluorescent protein (GFP). The ratio of EdU/Ki67/GFP triple-positive cells to EdU/GFP double-positive ones was determined.
Quantitative analysis of neuronal migration
Distribution of GFP-positive cells in the coronal sections were quantified by calculation of the number of labelled cells in each region of the brain slices.21
Statistical analysis
Results are presented as mean±SEM. Two-sided Student’s t-test was performed to compare the means between two groups. When the means of three groups were to be compared, one-way analysis of variance (ANOVA) with post-hoc Tukey’s honestly significant difference calculator test was used. Statistics were calculated using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan).22 P<0.05 was considered significant.
Results
Clinical findings
The subject of this study was a 15-year-old boy, the second son of unrelated and healthy Japanese parents. He was born at 36 weeks’ gestation with a birth weight of 2915 g (82th percentile), length of 48 cm (74th percentile) and head circumference of 35 cm (97th percentile). He was admitted to a neonatal intensive care unit because of multiple congenital anomalies, including ventricular septal defect, tracheomalacia and eventration of diaphragm. He gradually presented with respiratory distress and was intubated 40 days after birth with a subsequent tracheotomy performed at 3 months of age. At 7 months of age he was diagnosed with stage 4S neuroblastoma originating from the sympathetic nerve ganglia in his abdomen with lower limb metastasis,23 and attained remission with extirpative surgery and chemotherapy. He presented with postnatal growth retardation and megalencephaly, with a height of 146 cm (−3.4 SD), body weight of 30 kg (−2.6 SD) and head circumference of 62 cm (+5.9 SD) at 15 years of age. He had facial dysmorphic features of prominent forehead, thick and laterally extended eyebrow, posteriorly rotated ear, epicanthus, hypertelorism, wide and depressed nasal bridge, wide nasal base, upturned nasal tip, long philtrum, square face, and high-arched palate (figure 1A,B). He had postaxial polydactyly of the hands and feet (figure 1C). His global development was delayed, with a total developmental quotient (Kinder Infant Developmental Scale24) of 12 at 6 years of age. Developmental regression was not observed, and he was able to walk alone and speak several single words at the latest visit. Brain MRI showed ventriculomegaly and hypoplastic corpus callosum, but no cortical dysplasia like polymicrogyria (figure 1D–F). He developed generalised epilepsy at 10 years of age, which has been controlled by administration of zonisamide.
Clinical features of the patient (A and B). Distinctive facial features were observed, including prominent forehead, thick and laterally extended eyebrow, posteriorly rotated ear, epicanthus, hypertelorism, wide and depressed nasal bridge, wide nasal base, upturned nasal tip, long philtrum, and square face. (C) X-ray image showing postaxial polydactyly. (D and E) Axial slices of T2-weighted images acquired at 15 years show ventriculomegaly, but no polymicrogyria. (F) Sagittal slice of T2-weighted image taken at 17 months of age showing hypoplastic corpus callosum. The patient’s parents gave written consent for publication of the photographs.
Genetic findings
To identify the causative mutation underlying the patient’s phenotype, we performed trio-based, whole-exome sequence (figure 2A,B). Total reads by exome sequencing ranged from 54.0M to 65.8M reads (8.1G–9.9G bases), and the mean depth at the target region was 81.8× to 100.4×. Following filtering, seven candidate genes remained: four de novo mutations, one associated with an autosomal recessive trait and two associated with X linked recessive traits (figure 2A). After prioritisation, a de novo heterozygous missense mutation (NM_005378.5; c.173C>T; p.Thr58Met) in the MYCN gene was identified as the top candidate, which was confirmed by Sanger sequencing (figure 2C). It is noteworthy that the missense mutation was not listed in any public databases of general population (eg, ExAC), although it was present in the COSMIC database, somatic mutations in human cancer. In addition, Thr58 residue is evolutionally conserved (figure 2D). Indeed, the mutation was predicted to be pathogenic and damaging by in silico analyses (Polyphen-2 score=1.000, SIFT score=0.02 and CADD Phred score=33).
Genetic analysis of the patient. (A) The data analysis algorithm used for filtering all single nucleotide variants identified using trio-based, whole-exome sequence, with the number of remaining variants after each filtering step. On filtering and prioritisation, a de novo heterozygous missense mutation (c.173C>T; p.Thr58Met) in MYCN was identified as the top candidate. MAF, minor allele frequency. (B) Pedigree analyses with the patient represented by a filled symbol. The genotype of the variant p.Thr58Met or wild-type (WT) in the MYCN gene is given. (C) Genomic DNA sequence chromatograms indicating the position of the c.173C>T mutation. (D) Phylogenic conservation was observed for the mutated amino acid residue, marked by an asterisk.
Developmental changes of Mycn expression
Analysis of MYCN mRNA expression during fetal human brain development found it was highly expressed in immature neuronal cells, but this expression decreased after terminal differentiation.7 Only fragmentary information is available, however, for mouse Mycn protein expression during brain development. Thus, to better understand the role of Mycn in neuronal development, we examined the temporal expression pattern of Mycn in whole-tissue extracts of mouse brains at various developmental stages. Mycn (~65 kDa) was highly expressed at E13.5, then gradually decreased to its lowest level by postnatal day (P) 15 (figure 3). Sept11, a cytoskeleton-related protein, was visualised as a loading control (figure 3). This result confirmed that the temporal expression pattern of Mycn protein in the mouse brain reflected the situation in humans, and suggests Mycn plays an important role in brain development.
Expression profile of Mycn protein during mouse brain development. Tissue extracts of whole-brain (E13.5, E15.5, E17.5, P0) and cerebral cortices (P3, P8, P15, P30) were subjected to western blotting with an anti-N-Myc antibody. Anti-Sept11 was used as a loading control.
Effect of the c.173C>T mutation on Mycn stability and Ccnd1/2 expression
As the identified missense mutation c.173C>T occurs at T58, phosphorylation of which is crucial for MYCN degradation in neuronal precursor cells,25 we hypothesised the mutant MYCN-T58M protein accumulates due to being constitutively non-phosphorylatable. To test this possibility, we overexpressed MYCN-T58M and wild-type MYCN in HEK293T cells. We also overexpressed MYCN-P44L, which is the most common missense mutation of MYCN according to the COSMIC cancer cell database. While the expression of total MYCN was comparable between HEK293T cells expressing wild-type MYCN and MYCN-T58M, the antibody against T58-phosphorylated MYC showed negligible T58 phosphorylation in MYCN-T58M compared with wild-type MYCN. In addition, we observed an increased level of phosphorylation at T58 and S62 in MYCN-P44L compared with wild-type (figure 4A; one-way ANOVA with post-hoc Tukey’s honestly significant difference calculator test; df=2, F=59.2, p<0.01 for p-T58/MYCN; df=2, F=8.42, p<0.05 for p-S62/MYCN). These results suggest the possibility that mutations affecting the phosphorylation state at T58 or S62 change the biological activity of MYCN.
Effects of the p.T58M mutation on Mycn stability and Ccnd expression in neuronal progenitor/stem cells. (A) Expression of MYCN, MYCN-T58M and MYCN-P44L in HEK293T cells. Lysates from cells expressing MYCN, MYCN-T58M or MYCN-P44L were subjected to sodium dodecyl sulfate-poly acrylaide gel (SDS-PAGE) followed by western blotting with anti-N-Myc, phospho-c-Myc (pThr58), phospho-n-Myc (pSer62) or Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH) antibodies. GAPDH immunoreactivity indicates equivalent loading. Antiphosphorylated-c-Myc (pThr58) antibody recognised wild-type (WT) MYCN, but not MYCN-T58M. In contrast, MYCN-P44L mutant was associated with significantly increased levels of MYCN forms phosphorylated at Thr58 and Ser62. Fold change of phosphorylated MYCN level with respect to MYCN expression level was normalised to MYCN-WT from three independent experiments. (B) Expression of Mycn and Mycn-T58M in neuronal progenitor/stem cells in the presence (left two lanes) or absence (right two lanes) of Fbw7. Quantification of Mycn levels as a proportion ofgreen fluorescent protein (GFP) levels averaged over three independent experiments. Relative densities of the bands were determined relative to one for wild-type. Electroporation with Mycn-T58M caused an accumulation of mutant protein compared with wild-type. (C) Western blotting of Mycn and Mycn-T58M in neuronal progenitor/stem cells. Quantification was performed as in (B). Mycn levels as a proportion of Gapdh levels averaged over six independent experiments. Mycn-T58M was expressed at levels much higher than wild-type at 72 hours. (D) Expression of Ccnd1 and Ccnd2 in neuronal progenitor/stem cells transfected with control pCAG (Empty), pCAG-Myc-Mycn (WT) or pCAG-Myc-Mycn-T58M, together with pCAG-EGFP. Quantification was done as in (B). Protein levels as a proportion of Gapdh levels averaged over six independent experiments. Significantly higher expression was observed in cells expressing Mycn-T58M at 72 hours. (B–D) Bar graphs, means and SEM are shown; *p<0.05, **p<0.01.
Next, to assess the pathophysiological significance of the mutation in brain development, we analysed the effect of the p.T58M mutation on protein stability in neuronal progenitor/stem cells. When pCAG-Myc-Mycn or pCAG-MYC-Mycn-T58M was electroporated into neurosphere cells with or without pCAG-Myc-Fbw7 (ubiquitin ligase16), Mycn-T58M was shown to be more stable than wild-type Mycn 24 hours after electroporation (two-sided Student’s t-test; p<0.01 for Fbw7+, p=0.045 for Fbw7−; statistical analysis was run separately for Fbw7 +and Fbw7−) (figure 4B). Degradation of Mycn was much more prominent when the ubiquitin-proteasome system was enhanced by coexpression of Fbw7 (figure 4B); thus, subsequent experiments were performed with Fbw7 coexpression. Seventy-two hours after electroporation, Mycn-T58M was still expressed at high levels, approximately eight times higher than wild-type Mycn (8.6±1.5 [mean±SEM], p<0.01) (figure 4C). These results support the hypothesis that the non-phosphorylatable c.173C>T mutation confers resistance to ubiquitin-proteasome-mediated degradation of Mycn.
We next asked whether the p.T58M mutation affects neuronal cell proliferation. To this end, we prepared neuronal progenitor/stem cells and assessed the expression levels of Ccnd1 and Ccnd2, Mycn target gene products that promote cellular division and proliferation.26–28 As shown in figure 4D, overexpression of both Mycn-T58M and wild-type Mycn induced expression of Ccnd1 and Ccnd2, but significantly higher expression was observed in cells expressing Mycn-T58M (one-way ANOVA with post-hoc Tukey’s honestly significant difference calculator test; df=2, F=19.5, p<0.01 for Ccnd1; df=2, F=43.9, p<0.01 for Ccnd2; expression level was normalised to Gapdh). We cotransfected GFP to check the proportion of transfected cells because excess expression of Mycn was reported to cause apoptosis.29 30 We confirmed that the same amount of GFP is expressed at 24 hours after transfection (online supplementary figure 1). However, overexpression of both wild-type Mycn and Mycn-T58M decreased expression level of GFP, and cells with Mycn-T58M were strongly affected than wild-type Mycn 72 hours after transfection (figure 4D), indicating putative cell death induced by overexpression of wild-type Mycn or Mycn-T58M. Therefore, overexpression of wild-type Mycn or Mycn-T58M induces bidirectional effects on neuronal cell proliferation and apoptosis, and Mycn-T58M showed strong effects on both cellular events.
Supplemental material
Effect of Mycn expression on neuronal progenitor/stem cell proliferation and migration
Amplification of MYCN is found in about 25% of cases of neuroblastoma, which originates from undifferentiated neuronal crest cells. Megalencephaly associated with CCND2 gene abnormalities is thought to be caused by dysregulation of neuronal progenitor cell cycle, which increases their cell numbers.11 So far, we have shown that Mycn-T58M induced Ccnd1 and Ccnd2 expression on its protein stability. We therefore investigated whether the increased expression level of Mycn affects corticogenesis by assessing its effect on neurogenesis in mouse neuronal progenitor/stem cells in the ventricular (VZ) or subventricular/intermediate (SVZ/IZ) zones. Cells were triple-stained for EdU, GFP and Ki67, a marker for all active phases of the cell cycle except the quiescent G0 state. Cells that remained proliferating after EdU incorporation were identified as EdU/Ki67 double-positive, while neurons that differentiated after EdU incorporation were EdU-positive but Ki67-negative (figure 5A). Wild-type Mycn promoted cell cycle progression in the SVZ/IZ (figure 5B) (two-sided Student’s t-test; p<0.01; Empty, n=5; wild-type, n=4). Wild-type Mycn also tended to increase the number of EdU/Ki67 double-positive cells in the VZ, but this increase was not statistically significant.
Effect of Mycn overexpression on cell cycle and migration in neuronal progenitor/stem cells. (A) E14.0 cortices were co-electroporated with pCAG-EGFP together with pCAG-Myc (Empty) or pCAG-Myc-Mycn-WT (WT) and coronal sections immunostained for green fluorescent protein (GFP) (green), EdU (red) and Ki67 (blue). GFP/EdU/Ki67 triple-positive cells are indicated by arrowheads. Scale bar, 10 µm. (B) Quantification of GFP/EdU/Ki67 triple-positive cells relative to GFP/EdU double-positive ones in (A). Wild-type Mycn promoted cell cycle progression in the SVZ/IZ. (C) In utero transfection was done as in (A), followed by fixation at P0. Coronal sections of samples transfected with Empty (control) or wild-type Mycn were immunostained for GFP. (D) Quantification of the distribution of transfected neurons in distinct parts of the cerebral cortex for each condition shown in (C). (B, D) Bar graphs, means and SEM are shown; *p<0.05, **p<0.01. EdU, 5-ethynyl-2’-deoxyuridine; SVZ/IZ, subventricular/intermediate zones.
We next sought to study the consequences of overexpression of wild-type Mycn for neuronal positioning. We electroporated Mycn constructs in combination with enhanced green fluorescent protein (EGFP) reporter construct into the VZ of E14.5 mouse neocortices and analysed at P0. In P0 brain sections, we observed that neurons electroporated with the empty vector reached the superficial layers II–IV of the cortical plate (figure 5C). However, electroporation of wild-type Mycn induced arrest of cells within layers V–VI (two-sided Student’s t-test; p<0.05 for layers II–IV and layers V–VI; Empty, n=3; WT, n=3). In addition, Mycn-T58M affected cell proliferation and migration equally to wild-type Mycn (data not shown).
Discussion
To our knowledge, this is the first report of a germline pathological gain-of-function missense mutation in MYCN (c.173C>T; p.Thr58Met) that may cause a megalencephaly syndrome with ID, ventriculomegaly, hypoplastic corpus callosum, distinctive facies, postaxial polydactyly and neuroblastoma. We show that MYCN-T58M functions in an unphosphorylated form with increased stability, leading to excess accumulation of the protein. Moreover, Mycn-T58M promoted the expression of Ccnd1 and Ccnd2 more significantly than wild-type Mycn, strongly suggesting that neuronal progenitor cells expressing Mycn-T58M are more proliferative than control cells in vivo. Ala replacement at Thr58 in Mycn has been shown to have transactivation capacity and enhanced stability, equivalent to the T58M substitution.25 Furthermore, heterozygous mutations that stabilise CCND2 cause MPPH.11 In this context, comparison of clinical features of the present patient with those of patients with MPPH with CCND2 mutations revealed common phenotypes, including ID, megalencephaly, ventriculomegaly and postaxial polydactyly (table 1). We assume that aberrant MYCN-CCND signalling is responsible for these shared clinical features.
Comparison of clinical features in the present patient, and patients with Feingold syndrome or MPPH
Physiological MYCN activity is controlled via its regulated expression through sequential phosphorylation at S62 and T58. Phosphorylation at S62 by CDK1 via the Ras-Erk pathway confers resistance to proteolysis and thus maintains the transcriptional activity.14 This phosphorylation primes the subsequent phosphorylation at T58 by GSK3β via the PI3K-AKT pathway.13 As a result, cells become highly sensitive to ubiquitination by Fbw7.16 31 On the other hand, since MYCN-T58M does not undergo the second phosphorylation, it is resistant to degradation by the ubiquitin-proteasome system, resulting in accumulation of the mutant protein followed by increased CCND1/2 expression.
Like other reported gene mutations causing MPPH, the p.T58M mutation in MYCN induces CCND2 expression and causes some similar clinical phenotypes. There are, however, some distinguishing features observed in the present patient. In one study, all reported patients with MPPH with pathogenic mutations in CCND2, as well as in AKT3, PIK3R2 and PIK3CA, were complicated with polymicrogyria.32 Whereas brain MRI did not detect polymicrogyria in the present case, abnormal cortical architecture could be seen by histopathological analysis, based on the result that expression of Mycn-T58M induced neuronal migration delay in fetal mouse neocortices. Additionally, the corpus callosum of the present patient was considerably thinner than that of patients with MPPH. These discrepancies suggest the presence of yet unidentified MYCN signalling pathways that are independently dysregulated in MPPH and the present case. Further analyses are required to clarify the molecular mechanisms underlying the clinical diversity.
Thirty-seven loss-of-function abnormalities in MYCN have been listed in the HGMD to date.33 These gene abnormalities cause Feingold syndrome and can be broken down into 7 missense, 6 nonsense and 16 frameshift mutations and 8 gross deletions. Given that all the missense mutations were positioned in the basic-helix-loop-helix-leucine zipper domain, transcriptional activity of MYCN should be attenuated through impaired DNA-binding.8 34 35 Indeed, Feingold syndrome is thought to be caused by MYCN loss of function that results in a variety of symptoms, such as microcephaly and short stature.8 36 The HGMD also includes three gross insertions that exert pathological gain-of-function effects leading to tumours such as neuroblastoma and Wilms’ tumour.33 Notably, two of the three cases with gross insertions showed additional clinical features other than tumours. In addition to developmental delay, distinctive facial features are common symptoms, one showed postaxial polydactyly and the other presented ventriculomegaly, agenesis of the corpus callosum and ventricular septal defect.37–39
Our data indicate that c.173C>T (p.Thr58Met) in MYCN exerts gain-of-function effects via increased protein stability and increased neuronal proliferation. Since the patient in this study presented phenotypes similar to those of patients with MYCN duplication,37 39 we assume that expression of both wild-type Mycn and Mycn-T58M could facilitate neuronal progenitor cell proliferation. Thus, the sharp contrast in clinical phenotypes between the present case and those diagnosed with Feingold syndrome is likely to be derived from upregulation of the MYCN-CCND-mediated transcription pathway in the former and downregulation of the pathway in the latter (table 1).
Given that Mycn-T58M is thought to have gain-of-function effects, overexpression of wild-type Mycn should mimic the mutation phenotypes in developing cerebral cortex. Indeed, cortical neuron proliferation was facilitated by exogenous Mycn in the SVZ and IZ, but not in the VZ where apical progenitor cells (radial glia) accumulate. This may correlate with the dominant distribution of CCND2-positive cells in the SVZ rather than the VZ.40 Neuronal progenitor cells in the SVZ and IZ are categorised as basal progenitors essential for the human-specific development of neocortex.41 Thus, dysregulated MYCN-CCND2 function is possibly involved in megalencephaly and microcephaly in human diseases. The electroporation method used in this study results in different expression levels of protein among cells, dependent on the amount of incorporated expression vector. Our experiments showed that introduction of high levels of Mycn into neuronal progenitor cells appeared to cause cell death in vitro. Because of this non-uniformity of gene expression, cells overexpressing Mycn might have died and the result of our experiment appears not to reflect the accurate proliferative function of Mycn-T58M. We assume that subtle or moderate transcriptional activity by exogenous Mycn is enough to accelerate neurogenesis in the SVZ and IZ, since MYCN duplication also gives rise to similar clinical features.37 39 From these results, we consider that the expression level of Mycn may determine cell fate, proliferation or cell death; low to moderate Mycn expression promotes cell proliferation, whereas high expression results in cell cycle arrest followed by cell death. Therefore, gain-of-function effects of the MYCN-T58M mutant with increased stability may be enough to accelerate neurogenesis equivalent to that of duplication of the entire MYCN and cause megalencephaly via excess production of neuronal cells during corticogenesis.
In summary, we identified a de novo gain-of-function heterozygous missense mutation in MYCN in a patient with a novel megalencephaly syndrome similar to, but distinct from, PI3K-AKT-CCND2-related MPPH. This finding, and our subsequent functional analysis of the mutation, provides new insight into the critical role of MYCN in brain development.
Acknowledgments
The authors thank the patient and his parents for participating in this study.
References
Footnotes
Contributors SS was responsible for the concept and design of the study. KoK, KN and SS drafted the main manuscript. KoK, FM, NH, YN, HI, IH, KN and SS analysed and interpreted the data. KoK, YK, HO and SS contributed clinical data. AH, NO, MK, TT, YK, KeK and YT revised the manuscript and made comments on the structure, details and grammar of the article.
Funding This study was partially supported by the JSPS KAKENHI (grant number JP16K15530) (SS) and by the Program for an Integrated Database of Clinical and Genomic Information from the Japanese Agency for Medical Research and Development, AMED (SS).
Competing interests None declared.
Ethics approval This study was approved by the Ethical Committee for the Study of Human Gene Analysis at Nagoya City University Graduate School of Medical Sciences.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Obtained.