Heterozygous loss of function of NR4A2 is associated with intellectual deficiency, rolandic epilepsy, and language impairment.

Recognition of a de novo mutation in NR4A2 associated with a neurodevelopmental phenotype reinforces its role in 2q23q24 microdeletion syndrome. Using the proband WES data and the probability of loss-of-function intolerance index (pLi) set at 1.0 (highest intolerance constraint), we could target NR4A2 as the candidate gene in this patient.

BACKGROUND

NR4A2 haploinsufficiency due to whole gene deletion was recently associated with neurodevelopmental and language delay phenotypes. We report a heterozygous loss‐of‐function variant in NR4A2 [nuclear receptor subfamily 4 group A2, OMIM *601828] in a patient with epilepsy, language impairment, and intellectual deficiency. This finding confirms that NR4A2 is responsible for the features associated with the 2q23q24 microdeletion syndrome.

NR4A2 haploinsufficiency due to whole gene deletion was recently associated with neurodevelopmental and language delay phenotypes by Reuter et al1 and Lévy et al.2 This finding strongly suggests that NR4A2 is the critical gene for 2q24.1 microdeletion syndrome.

CASE PRESENTATION

Herein, we report a heterozygous loss‐of‐function variant in NR4A2 [nuclear receptor subfamily 4 group A2, OMIM *601828] in a patient with epilepsy, language impairment, and intellectual deficiency. The individual is the first son of nonconsanguineous Caucasian parents, who was born after an uneventful pregnancy and labor. During infancy, he presented poor feeding and gastrointestinal symptoms such as gagging due regurgitation and colic. The boy had normal motor development and growth parameters, but his speech development was delayed. Dysmorphic features were absent. At age of 5 years, he presented seizures during sleep. Electroencephalogram (EEG) background activity was normal; interictal activity was characterized by high amplitude bilateral sharp waves paroxysms in the centroparietotemporal region, with marked frequency increase during drowsiness and sleep. This pattern was suggestive of rolandic epilepsy but did not characterize electrical status epilepticus during slow‐wave sleep (ESESS). Sporadic motor focal seizures persisted for 2 years, usually during sleep, and are currently under control with an association with clobazam and sulthiame. His EEG did not change along these years. Brain MRI was normal. Formal cognitive testing at 9 years of age with WISC‐IV (Wechsler Intelligence Scale for Children) revealed results below −2 SD. Verbal comprehension and operational memory were particularly impaired, with a calculated IQ of 63 (1st percentile) and 52 (0.1th percentile), respectively. He currently presents learning disability and in inclusive education at school.

Whole exome sequencing revealed a frameshift variant in NR4A2 (c.326_327insA or c.326dupA, ENST00000339562) which is predicted to cause a premature stop codon [p.(Ser110Valfs*2)] leading to either a nonsense‐mediated decay or a shortened nonfunctional protein. Sanger sequencing confirmed its presence in the patient, and additional study of the parents showed that it occurred as a de novo event Figure 1. No other potentially pathogenic variants in genes previously associated with intellectual deficiency or rolandic epilepsy, such as GRIN2A and KCNB1, were found.

Figure 1

Electropherogram showing the presence of heterozygous c.326dupA variant in patient and its absence in parents, characterizing a de novo mutation

DISCUSSION

To our knowledge, this is the first time a point mutation leading to loss of function in NR4A2 is reported. Our patient's phenotype is similar to the previously described in individuals with de novo microdeletions encompassing NR4A2 at the 2q24.1 region and whose clinical manifestations were mainly characterized by language and cognitive delay.1, 2, 3, 4, 5, 6 Our patient also presented focal motor seizures in the rolandic epilepsy spectrum, which expands the clinical manifestations related to NR4A2 haploinsufficiency and reinforces the predominance of language impairment associated with this condition.

Barge‐Schaapveld et al3 and Shimojima et al6 suggested previously that GPD2 haploinsufficiency was the most likely to be responsible for the 2q23q24 microdeletion phenotype. However, GPD2 probability of loss‐of‐function intolerance (pLi) is zero and over 100 loss‐of‐function variants are reported in ExAC,on the other hand, NR4A2 has a pLi = 1 and no loss‐of‐function variants in ExAC.7 This finding reinforces the probability of NR4A2 being the main gene responsible for the phenotype of 2q23q24 microdeletion. It also demonstrates the utility of pLi to funnel down candidate genes for a microdeletion phenotype.

Seizures pattern and EEG of the patient are within the spectrum of rolandic epilepsy, in which language and behavior abnormalities are also common. In the more severe end of this continuum, Landau‐Kleffner syndrome and ESESS, developmental delay, language regression, and behavioral disorders are commonly observed.8 Pathogenic variants in other genes, such as GRIN2A and KCNB1, have been previously reported associated with rolandic epilepsy spectrum9, 10 reinforcing this neurocognitive endophenotype.

Our finding confirms that NR4A2 is responsible for most, if not all, clinical features associated with the 2q23q24 microdeletion syndrome, furthermore, reinforces its implication in language impairment, and adds epilepsy to the phenotype.

CONFLICT OF INTEREST

None declared.

AUTHOR CONTRIBUTIONS

LLPR: was responsible for data collection and manuscript writing. FPM, LAC, and MDOR: were involved in whole genome sequencing analysis and contributed in manuscript writing. LPBS: performed epilepsy investigation. ELF: is the laboratory senior scientist and was responsible for NGS and Sanger sequencing. JPK: is the senior bioinformatician and was involved in variants filtering and prioritization. FK: was responsible for patient's care, genetic analysis and phenotype characterization as well as manuscript edition.