Background DYRK1A plays different functions during development, with an important role in controlling brain growth through neuronal proliferation and neurogenesis. It is expressed in a gene dosage dependent manner since dyrk1a haploinsufficiency induces a reduced brain size in mice, and DYRK1A overexpression is the candidate gene for intellectual disability (ID) and microcephaly in Down syndrome. We have identified a 69 kb deletion including the 5′ region of the DYRK1A gene in a patient with growth retardation, primary microcephaly, facial dysmorphism, seizures, ataxic gait, absent speech and ID. Because four patients previously reported with intragenic DYRK1A rearrangements or 21q22 microdeletions including only DYRK1A presented with overlapping phenotypes, we hypothesised that DYRK1A mutations could be responsible for syndromic ID with severe microcephaly and epilepsy.
Methods The DYRK1A gene was studied by direct sequencing and quantitative PCR in a cohort of 105 patients with ID and at least two symptoms from the Angelman syndrome spectrum (microcephaly < −2.5 SD, ataxic gait, seizures and speech delay).
Results We identified a de novo frameshift mutation (c.290_291delCT; p.Ser97Cysfs*98) in a patient with growth retardation, primary severe microcephaly, delayed language, ID, and seizures.
Conclusion The identification of a truncating mutation in a patient with ID, severe microcephaly, epilepsy, and growth retardation, combined with its dual function in regulating the neural proliferation/neuronal differentiation, adds DYRK1A to the list of genes responsible for such a phenotype. ID, microcephaly, epilepsy, and language delay are the more specific features associated with DYRK1A abnormalities. DYRK1A studies should be discussed in patients presenting such a phenotype.
- Clinical genetics
- Molecular genetics
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
DYRK1A/MNB is a protein kinase that belongs to the highly conserved dual-specificity tyrosine phosphorylation-regulated kinase (DYRK) family, which has been shown to be expressed in a gene dosage dependent manner. It participates in multiple biological pathways, and plays differential functions during development and in the adult brain.1 Orthologous genes have been cloned independently in various organisms and named Minibrain (Mnb) or DYRK1A. DYRK1A catalyses its intrinsic autophosphorylation on tyrosine residues for self-activation without requiring any cofactor, as well as the phosphorylation of serine/threonine residues on exogenous substrates. It also localises in the Down syndrome critical region on human chromosome 21 and was identified as the candidate gene for diverse features of Down syndrome, including intellectual disability (ID), microcephaly, and the age-associated neurodegeneration.2 ,3 DYRK1A is increased 1.5-fold in a gene dosage-dependent manner in Down syndrome patients as well as in Down syndrome fetal brains.4 ,5 Many proteins have been identified as possible substrates and/or interacting proteins, including transcription factors such as CREB, FKHR, GLI1, NFAT, STAT3, APP, tau, and α-synuclein, but the physiological substrates/interacting partners of DYRK1A in neural development remains mostly unknown.
However, DYRK1A has been shown to exhibit crucial functions during central nervous system (CNS) development, in regulating neural proliferation, neurogenesis, neuronal differentiation, cell death, and synaptic plasticity.6–11 DYRK1A also regulates cell death through facilitating apoptosis signal-regulating kinase 1 (ASK1)-mediated signalling events,12 and is also involved in the cell cycle and mitosis of non-neuronal cell lines, potentially fulfilling multiple actions in cell cycle regulation in different tissues such as cardiomyocytes.13
To date, four patients with intragenic DYRK1A rearrangements or overlapping 21q22 microdeletions including only DYRK1A have previously been reported. They present with ID, absent or delayed language, microcephaly, and epilepsy.16–18 A few other patients have been reported with larger rearrangements including DYRK1A and additional proximal genes, and also present with a similar phenotype.19 ,20 DYRK1A mutations have never been reported. The DYRK1A gene had been sequenced in a cohort of 150 epileptic patients but no mutation has been identified.18
We report the identification of DYRK1A as a new gene responsible for ID, severe microcephaly, epilepsy, and growth retardation, and discuss the pathophysiogical implication of this gene in this phenotype.
Patients and methods
The index case was referred to our unit for genetic investigations because of developmental delay. She was the first child from healthy non-consanguineous parents. She was born at 39 weeks of gestation with intrauterine growth retardation (length=42.5 cm (−4 SD), weight=1860 g (−3 SD)) and microcephaly (occipitofrontal circumference (OFC)=30 cm; −4.5 SD). During her first months of life (figure 1A), she presented with feeding difficulties. At 1 year of age, epileptic seizures occurred in a febrile context requiring valproate treatment. The electroencephalogram (EEG) features of the index case at 15 months of age revealed 6 Hz background activity and movement and eye artefacts (figure 1E), and persistent theta activity of 6–7 Hz mainly in the occipital region, and in non-rapid eye movements sleep stage 2, diffuse 5–10 s theta activity bursts in centrotemporal regions without any clinical signs. Her developmental skills were impaired: she presented with hypotonia, was able to sit at 15 months of age, and to walk independently at 3 years of age. Neurological examination showed evidence of an abnormal gait due to ataxia. At 3 years of age, cerebral MRI was normal. Standard chromosomal analysis and SNRPN (Small Nuclear Ribonucleoprotein Polypeptid N) methylation appeared normal, as well as standard metabolic screening. Oligonucleotide array comparative genomic hybridisation (CGH) 105K (Agilent Technologies, Santa Clara, California, USA) identified a 69 kb de novo deletion including the 5′ region of the DYRK1A gene (arr 21q22.13(37 644 501–37 713 641)×1 dn (hg19)) (figure 2).
At 4 years of age, she presented with growth retardation (−3 SD) and severe microcephaly (−6 SD), whereas she had gained weight (−2 SD). Facial dysmorphism included thick lips, bulbous nose, mild hypotelorism, micrognathia, prominent incisors, and large ears with thick helix (figure 1A–D). Her language was delayed, and she could only repeat syllables. The ataxia had disappeared and she no longer had an abnormal gait. There was evidence of behavioural abnormalities with periods of inappropriate laughs and hyperactivity.
Cohort of patients
A cohort of 105 patients was recruited from the institutions of two French experts on Angelman syndrome (AS) when they met the following criteria: ID and at least two of the following additional symptoms: microcephaly < −2.5 SD, ataxic gait, and seizures. All parents had given their informed consent for research analyses. Standard karyotype (105/105 cases), high resolution array CGH (30/105 cases), methylation at the SNRPN locus (105/105 cases) or UBE3A sequencing analyses (75/105 cases) were normal.
DYRK1A DNA sequencing analysis
For mutation analysis of the gene DYRK1A, the coding region and the flanking intron-exon boundaries were PCR amplified with primers based on the genomic DNA: NM_001396.3 (primer sequences available on request). PCR products were purified with ExoSAP-IT (USB Corporation, Cleveland, Ohio, USA) followed by bidirectional sequencing with forward and reverse primers with BigDye Terminator v3.1 Cycle Sequencing Kit on a fluorescent DNA sequencer (ABI-Prism 3100 sequencer, Applied Biosystem, Foster City, California, USA). Sequences were analysed with SeqScape v2.7 software (Applied Biosystem, Foster City, California, USA).
DYRK1A DNA quantification
DYRK1A DNA quantification was performed by DNA quantitative PCR (qPCR) in the 75 patients who did not have array CGH analysis.
DYRK1A DNA qPCR was performed using LightCycler 480 SYBR Green I Master Mix. The HRPP1 (Ribonuclease P, RNA Component H1) gene was used for normalisation. DYRK1A exons 3 and 9 were amplified with primers selected within regions of unique sequence and designed with Primer3Plus (primers available on request).21 Relative quantification was performed on LightCycler 480 (Roche Applied Science, Indianapolis, Indiana, USA) following the manufacture's protocol. All samples were run in triplicate. Data were analysed with the LightCycler 480 Software (Roche Applied Science).
Cohort of patients
Among the cohort of 105 patients, 44 were males and 61 were females. The mean age was 11 years. They presented with ID (105/105 cases), microcephaly (78/105 cases), seizures (58/105 cases), and/or ataxic gait (47/105 cases). No family history of similar clinical phenotype was found.
DYRK1A sequencing analysis
DYRK1A direct DNA sequencing identified a causal de novo heterozygous frameshift mutation (c.290_291delCT; Ser97Cysfs*98) in exon 3 in individual D12 (figure 2). The patient was a 14-year-old girl with a history of intrauterine growth retardation (birth measurements: length −4 SD, weight −2 SD, and OFC −4 SD) and feeding difficulties since the first months of life. Gastro-oesophageal reflux was first diagnosed. From the age of 18 months to 6 years, she presented with episodes of vomiting, anorexia and dehydration. Complete metabolic and digestive screening remained negative. Febrile seizures were first noted at 18 months of age and could be of myoclonic type. Atonic seizures followed by tonic-clonic generalised seizures appeared at 3 years of age. Treatment was refused by the parents until the finding of atonic seizure, sometimes followed by postictal left brachiofacial hemiparesia at 8 years of age. EEG analysis showed 2 s bursts of generalised 4 Hz slow waves mainly in the anterior region with temporal and posterior spikes (figure 1F). Generalised myoclonic seizures could accompany these bursts. Cerebral MRI revealed diffuse cortical atrophy with enlarged ventricles. At 14 years of age, she presented with ID, severe microcephaly (length −2 SD, weight −2 SD, and OFC −6 SD), severe speech delay (only a few words at 14 years of age), and facial dysmorphism (thick lower lip, mild hypotelorism and hypoplastic ear lobes). Neurological examination showed evidence of hand stereotypies.
This teenager was transferred to a centre for special needs. The neuropsychological assessment using the Weschler Preschool and Primary Scale of Intelligence, third edition (WPPSI-III) showed intellectual disabilities with a cognitive skill level which was equivalent to children between 4 years 7 months and 6 years 2 months of age. Her non-verbal aptitudes were higher than her verbal cognitive aptitudes. She also showed attention disabilities with impulsiveness when she had to make a choice and lower processing speed.
DNA sequencing also revealed an intronic heterozygous variation (c.208-13C>G) for patient M66. It was thought to be non-pathogenic since no difference was found between mutant and reference sequence with Human Splicing Finder.22
Four known non-pathogenic single nucleotide polymorphisms were identified: rs928763 in 92/105 patients, rs55650427 in 32/105 patients, rs2835772 in 48/105 patients, and rs55854596 in 1/105 patients.
DYRK1A DNA quantification
DYRK1A DNA quantification of exons 3 and 9 identified no abnormalities.
Here, we further delineate the clinical spectrum of DYRK1A rearrangements and mutations, giving particular emphasis to the presence of microcephaly in the phenotype.
We first identified a sporadic 69 kb DYRK1A microdeletion in the index case who exhibited the association of ID, microcephaly, delayed language, ataxic gait, and epileptic seizures. Four patients with intragenic DYRK1A rearrangements or overlapping 21q22 microdeletions including only DYRK1A had been previously reported (table 1).16–18 These patients presented the association of ID, microcephaly, growth retardation, epileptic seizures and severe speech delay in all cases. Microcephaly was severe (£-4 SD) in 2/4 cases, and ataxic gait was described in 2/4 cases.
We therefore hypothesised that the DYRK1A gene could be one of the underlying causes of a consistent pattern of clinical features, that is, ID, growth retardation, microcephaly, epilepsy and ataxic gait, and searched for DYRK1A mutation or genomic rearrangement in a cohort of 105 patients with such a phenotype. Because patients with DYRK1A aberrations showed some of the characteristic features of AS, that is, ID, microcephaly, ataxic gait, and epilepsy, the cohort was collected from patients with normal methylation at the SNRPN locus (105/105 cases) and/or normal UBE3A sequencing analyses (75/105 cases). A de novo DYRK1A frameshift mutation was identified in a patient with developmental delay, severe microcephaly (−6 SD), epileptic seizures, hand stereotypies, and feeding difficulties during infancy (table 1). DYRK1A sequencing analysis in a previous cohort of 150 patients did not identify any mutation,18 but inclusion criteria were ID and/or epilepsy of unknown aetiology. Since microcephaly appears to be one of the predominant features associated with DYRK1A abnormalities but was not considered in the inclusion criteria, it could explain the absence of DYRK1A mutations in their cohort. Indeed, microcephaly was present at birth and worsened during childhood in all the cases, and was striking by its severity (−3 to −6 SD). The microcephaly in patients with DYRK1A abnormalities appears more severe than in AS and thus might be an important differential feature with AS. Epileptic seizures, language delay or absent speech, intrauterine growth retardation, and feeding problems during infancy also appeared to be constant features (table 1). Dysmorphic features were unspecific but underdeveloped ear lobes appeared frequently (figure 1D). The presence of an epileptic disease starting with febrile seizures and characterised by the presence of atonic and myoclonic seizures is classic in AS,23 but EEG analysis did not reveal a specific EEG pattern of AS (theta slow waves bursts in particular) in both of our patients. Some other important findings of AS, that is, facial manifestations, clinical courses of epilepsy, and behaviour with smiling, were also not observed in patients with DYRK1A abnormalities.
However, DYRK1A cannot be considered as a predominant cause of such a phenotype, since only 1/105 patients exhibits a DYRK1A mutation. When microcephaly, epilepsy, and language delay/absent speech are considered as mandatory features, a DYRK1A mutation is found in 1/70 patients (1.4%). These results are of importance to further delineate which patients should be studied for this gene. Thus, the DYRK1A analysis could be mainly discussed when patients present with the association of ID, primary microcephaly, and epilepsy. Two other signs seemed specific: intrauterine and postnatal growth retardation and feeding difficulties during infancy. Such a phenotype also overlaps with monogenic disorders due to SLC9A6, MECP2 (Rett syndrome), CDKL5, FOXG1, ZEB2 (Mowat-Wilson syndrome), and TCF4 (Pitt-Hopkins syndrome) mutations, and chromosomal deletion such as 1p36 subtelomeric deletion and 2q23 deletion.24–27
Symptoms of the patients with DYRK1A mutations are likely to be due to DYRK1A haploinsufficiency, since it has been shown to be responsible for some phenotypic features in animal models, including microcephaly and intrauterine and postnatal growth retardation.14 ,15
Indeed, DYRK1A protein expression is correlated with DYRK1A copy number. Brain size is affected in correlation with DYRK1A copy number, mainly the hippocampal region, whereas cell density is inversely correlated with DYRK1A copy number.28 Relevant functions of DYRK1A in neurodevelopment had been deduced from the function of its Drosophila orthologue: minibrain gene (mnb). The mnb gene mutations resulted in brain size reduction and visual and olfactory behavioural deficiencies.15 Mutant Drosophila brain size was reduced in a region specific manner, mainly the optic lobes and central brain hemisphere. Neuroblasts were abnormally spaced and fewer than in the wild type Drosophila, but without alterations in neuronal architecture. DYRK1A has also been suspected as a candidate gene for microcephaly regarding the mouse phenotype. Indeed, homozygous null DYRK1A−/− presented delayed general growth and decreased brain size, with death during gestation.14 Heterozygous DYRK1A−/+ mice showed decreased neonatal viability, reduced body size from birth to adulthood, as well as reduced brain size and developmental delay.14 The brain size appears decreased in a region-specific manner, although the cytoarchitecture and neuronal components in most areas are not altered, suggesting that DYRK1A does not affect neuronal proliferation/differentiation in all CNS structures, despite a wide expression in the developing CNS.
The reduced brain size seems to be caused by a reduction in the total number of neurones and afferent fibres14 ,29 because DYRK1A is believed to have an important role in controlling brain growth through neuronal proliferation.28 DYRK1A inhibits the proliferation of neuronal progenitors via its kinase function and interaction with cycline kinase inhibitor (p27KIP1), resulting in cell cycle exit in neurones. DYRK1A loss of function results in an increased apoptosis of neuronal progenitors and finally to a reduced number of neurones. Neurogenesis has indeed typically been shown to be affected in ID disorders comprising primary microcephaly.30
The crucial roles of DYRK1A in the regulation of neurogenesis exhibit a dual function, regulating not only neuronal proliferation but also neuronal differentiation. DYRK1A inhibits the initiation of neuronal differentiation by suppressing NOTCH signalling and has also been implicated in late neuronal differentiation by having a possible role in dendritogenesis, regulating actin dynamics, and synaptogenesis by involvement in synaptic vesicle recycling.7 ,10 Such pathogenic roles have previously been implicated in many epileptic encephalopathies.31
Recently Mnb/Dyrk1a was found to regulate food intake in Drosophila and in mice via the regulation of neuroendocrine factor.32 An increase in Mnb/Dyrk1a expression resulted in an increase in food intake. DYRK1A haploinsufficiency led to weight and length retardation in both animal models and patients with DYRK1A abnormalities. Therefore it should be hypothesised that Mnb/DYRK1A haploinsufficiency could result in a decrease in food intake that could explain the important feeding difficulties presented by patients in childhood (4/4 cases).
In conclusion, the identification of an additional DYRK1A microdeletion and the first truncating mutation in an affected patient provides evidence that DYRK1A is a gene responsible for syndromic ID with microcephaly, ataxic gait, epilepsy, and growth retardation, and further delineates the phenotypic spectrum of patients with DYRK1A abnormalities. We suggest that DYRK1A should be added to the list of genes responsible for syndromic ID associated with severe microcephaly, growth retardation, and epilepsy.
Primer designing tool: (http://www.primer3plus.com); Control DNA SNP databases on dbSNP (http://www.ncbi.nlm.nih.gov/projects/SNP/), NHLB1 Exome Sequencing Project Exome Variant Server (http://evs.gs.washington.edu/EVS/), OMIM: (http://omim.org) and intronic variant pathogenicity assessment on Human Splicing Finder: (http://www.umd.be/HSF/).
DYRK1A reference sequence used was NM_001396.3. Protein sequence was NP_001387.2.
The authors thank the patients and their families for their contribution. The authors also thank the Regional Council of Burgundy, the Burgundy University and the Dijon University Hospital for their financial support.
Contributors Patient recruitment and phenotype: CT-R, AM, LF, J-BC, AM, PM, LL, CP, ML.
Competing interests None.
Patient consent Obtained.
Experimental analysis Jean-Benoît Courcet, Estelle Lopez, Nadège Gigot, Clémence Ragon, Julien Thevenon, Laurence Duplomb.
All the authors have participated in writing and reviewing the manuscript.
Provenance and peer review Not commissioned; externally peer reviewed.